CN100592777C - Information processing apparatus and method - Google Patents

Information processing apparatus and method Download PDF

Info

Publication number
CN100592777C
CN100592777C CN200510121796A CN200510121796A CN100592777C CN 100592777 C CN100592777 C CN 100592777C CN 200510121796 A CN200510121796 A CN 200510121796A CN 200510121796 A CN200510121796 A CN 200510121796A CN 100592777 C CN100592777 C CN 100592777C
Authority
CN
China
Prior art keywords
image
decoding
gop
processing unit
control
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Expired - Fee Related
Application number
CN200510121796A
Other languages
Chinese (zh)
Other versions
CN1893608A (en
Inventor
柴田正二郎
纲岛修二
小薮恭平
高村元嗣
柿田新次郎
金子光久
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Sony Corp
Original Assignee
Sony Corp
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Sony Corp filed Critical Sony Corp
Publication of CN1893608A publication Critical patent/CN1893608A/en
Application granted granted Critical
Publication of CN100592777C publication Critical patent/CN100592777C/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Abstract

The present invention relates to an information processing apparatus and method, recording medium, and program. An information processing apparatus for decoding compression-coded video data includes at least one decoder decoding the compression-coded video data, a supply controller controlling the supply of the compression-coded video data to the decoder, and a controller controlling processing executed by the supply controller and the decoder. The controller determines the order of decoding processing performed by the decoder so that, among pictures contained in a decoding processing unit forthe decoding processing performed by the decoder, I-pictures and P-pictures are decoded before B-pictures. The controller selects pictures to be output from the decoder from among the pictures contained in the decoding processing unit on the basis of a playback speed instruction provided by a playback speed instruction unit.

Description

Information processor and method
The cross reference of related application
The present invention comprises the theme of the Japanese patent application JP2005-242161 that is involved in the Japanese patent application JP 2004-311521 that submitted in Japan Patent office on October 26th, 2004 and submits in Japan Patent office on August 24th, 2005, and its full content is incorporated herein with as a reference.
Technical field
The present invention relates to information processor and method, recording medium and program.Or rather, the present invention relates to wherein can to reset compressed encoding video data and need not to carry out information processor and method, recording medium and the program of complicated group memory control.
Background technology
As video compression technology, use moving image encoding group/Motion Picture Experts Group (MPEG) standard widely.The replay operations of carrying out with the flow data of MPEG method coding by decoding not only comprises the normal playback operation, and comprises quick replay operations or reverse reproduction operation.
For example, in the long image sets of MPEG (GOP), each GOP comprises 15 width of cloth images, by omit (B) image of predictive coding before in being entered into decoder two-wayly, can with *-3 to * 3 (symbol "-" is represented in speed with reverse execution replay operations, and also so application in the following description) carry out quick replay operations (for example, referring to the open No.8-98142 of Japanese Unexamined Patent Application).In addition, for each playback speed provides the table of the image information of indicating to show, so that can carry out replay operations with the different speed of as the table shows same sequence number.
Another kind of reproducing process is disclosed in the open No.8-56334 of Japanese Unexamined Patent Application.In this technology, obtain 10 frame compressed encoding signals off and on the interval of 5 frames.After all 10 frame compressed encoding signals of demodulation, it is alternately supplied with two decoders, each decoder five frame, and with its write memory then of decoding.Memory reads every one frame so that can export * 2 replay signal.
Mpeg stream is suitable for the replay operations of forward most.For example, in the long GOP of MPEG, the coded sequence of 15 width of cloth images, promptly, form (I) image of the interior coding of a GOP, (P) image of predictive coding and B image are I (2), B (0), B (1), P (5), B (3), B (4), P (8), B (6), B (7), P (11), B (9), B (10), P (14), B (12) and B (13) (order of numeral when the GOP image rearranges in DISPLAY ORDER according to coded sequence in the bracket).The P image according to one the preceding I image or P image predict, and the B image according to one the preceding I image or P image and with two of coded sequence I image or P image are predicted the preceding.Therefore,, can decode to P image and two of B images by keeping two groups of reference picture store, thus the service efficiency of raising memory.
For mpeg stream, in the open No.10-150635 of Japanese Unexamined Patent Application, following technology has been proposed with the most suitable forward playback operation of reverse reproduction.A plurality of decoders are used to carry out decoding processing, and have directly increased GOP before the GOP that will decode, and two GOP are offered decoder.Even this feasible reference picture that in the reverse reproduction operation, also can correctly use the image that to reset.
In the open disclosed technology of No.8-98142 of Japanese Unexamined Patent Application, although can realize with *-3 to * 3 quick replay operations by before being entered into decoder, omitting the B image, it is irregular that but the demonstration of image becomes at interval, thereby make not nature of display image.
Wherein use a plurality of decoder chips that are used to carry out decoding processing technology can be omitted in input phase and the B treatment of picture that shows do not merged mutually.By this merging, can realize that the quick replay operations of regular intervals (for example, utilizes two decoder chips to realize * 2, utilize three decoder chips to realize * 4, utilize four decoder chips to realize * 5, and utilize five decoder chips to realize * 7).Yet in this case, output handle during need the omission of I image or P image, the complexity that this has increased the control operation of baseband memory or has been used for selecting to want from decoded data the processing of data presented.Therefore, be difficult by using known technology dynamically to change playback speed to carry out display operation.
As mentioned above, by the form of the image information that will show of representing each playback speed is provided, just can realize replay operations as the friction speed of the same sequence number on the form.Yet in this processing, in list cell, switch playback speed, and when switch speed, need complicated control to come display image reposefully.
As mentioned above, mpeg stream is suitable for the forward playback operation most.In reverse reproduction operation with the input sequence of mpeg stream, when carrying out the reverse reproduction operation after GOP of decoding, need the frame memory of at least one GOP of storage (for example, 15 frames), and should reserve many reference image data groups, thereby need big memory span.
Summary of the invention
In view of this background technology, need finish quick replay operations, reverse reproduction operation and oppositely quick replay operations by the video data of decoding compression-coded, and need not to carry out complicated group memory control.
According to the embodiment of the invention, a kind of information processor is provided, comprising: at least one decoding device is used for the video data of decoding compression-coded; Apparatus for controlling of supply, the video data that is used to control compressed coding is supplied with decoding device; And control device, be used to control the processing of carrying out by apparatus for controlling of supply and decoding device.Control device is determined the order of the decoding processing carried out by decoding device, so that among the image that is contained in the codec processing unit that is used for the decoding processing carried out by decoding device, and decoding I image and P image before the B image.The playback speed instruction that provides according to the playback speed command device, control device are selected among the image that is included in codec processing unit will be from the image of decoding device output.
Decoding device can comprise: the decoding processing final controlling element is used to carry out decoding processing; And the image apparatus for controlling of supply, be used for control image is supplied with the decoding processing final controlling element.The image apparatus for controlling of supply can be under the control of control device, is being included among the image of codec processing unit, will select it as being supplied to the decoding processing final controlling element from the I image of decoding device output image and P image and B image.The image that the decoding processing final controlling element can be decoded and be supplied with from the image apparatus for controlling of supply also can export the image of selecting as the image that will export from decoding device under the control of control device.
A plurality of decoding devices can be provided.Information processor can also comprise output switch apparatus, is used to receive the not packed data from a plurality of decoding device outputs, and is used for optionally exporting the not packed data that is received.Control device can also be controlled the processing of being carried out by output switch apparatus.
Decoding device can comprise and be used to store the group memory of decoded picture, and this group memory has and is less than the group that is included in the picture number in the codec processing unit.Decoding device can by use be stored in the group memory in through the decoding image as the reference image, carry out decoding processing as required.
The picture number that can store of group memory is bigger by 2 than the number that is included in I image in the codec processing unit and P image.
The number of decoding device can be three, and each decoding device can comprise the group memory that can store 8 width of cloth images.
Control device can be controlled decoding device so that storing I image and P image for the I image of group memory and the fixed position of P image.
When the image of the stem of codec processing unit is I image among the decoded picture of decoding device output or P image, control device control decoding device in case the demonstration that the decoding that decoding device begins to decode begins regularly and decoding device begins the output decoder image illustrate timing each other phase deviation first pre-determine the image of number.
First pre-determines total big by 1 that number can be than I image and P image.
When the image in the codec processing unit stem is B image among the decoded picture of decoding device output, control device can control decoding device in case the demonstration output timing that the decoding that decoding device begins to decode begins regularly and decoding device begins the output decoder image each other phase deviation second pre-determine the image of number.
The second predetermined number can be than I image and P image total big by 2.
When the playback speed of playback speed command device indication and playback direction are normal and forward respectively, control device control decoding device in case the demonstration that the decoding that decoding device begins to decode begins regularly and decoding device begins the output decoder image export timing each other phase deviation first pre-determine the image of number.
The first predetermined number can be than I image and P image total big by 1.
When the playback speed of playback speed command device indication and playback direction when being respectively normal and reverse, control device control decoding device in case the demonstration that the decoding that decoding device begins to decode begins regularly and decoding device begins the output decoder image export timing each other phase deviation second pre-determine the image of number.
Second pre-determines total big by 2 that number can be than I image and P image.
Control device can determine that the timing of decoding device execution decoding processing is so that the timing ratio specific output of decoding device decoding B image is wanted piece image early through the timing of the B image of decoding.
The video data of compressed coding comprises GOP.When reception comprises the GOP of I image or P image, the number of described image sets deducts 3 values that obtain greater than the group number from the group memory, control device is divided into codec processing unit with image sets, each codec processing unit comprises having than deducting the number I image or the P image still less of 2 values that obtain by the group number from the group memory, and the control apparatus for controlling of supply is to supply with decoding device with the video data of the compressed coding of codec processing unit.
Control device can form codec processing unit so that be included in the I image divided in the codec processing unit or the sum of P image can not become than formerly group or the total big predetermined number of the I image in formerly the codec processing unit or P image or more on the time on the time that is included in image.
The video data of compressed coding can comprise GOP.When reception comprises a GOP of I image or P image, the number of described image deducts 3 values that obtain less than the group number from the group memory, go up control device detection time continuously in the structure of the 2nd GOP of a GOP, and, if the sum that is included in I image among a GOP and the 2nd GOP or P image is less than deducting 2 values that obtain by the group number from the group memory, control device merges a GOP and the 2nd GOP forming codec processing unit so, and the control apparatus for controlling of supply is so that supply with decoding device with the video data of the compressed encoding of codec processing unit.
Control device can form codec processing unit, can not become than the total big predetermined number of I image in formerly the group on the time that is included in image or P image or more so that be included in the sum of the I image that merges in the codec processing unit or P image.
The video data of compressed coding can comprise GOP.Control device can detect the structure of a GOP and the structure of the 2nd GOP in a GOP continuous in time, if and the sum that is included in I image among a GOP and the 2nd GOP or P image is less than by deducting two times of 3 values that obtain from the group number of group memory, control device merges a GOP and the 2nd GOP so, divide the GOP merged then to form first codec processing unit and second codec processing unit, each codec processing unit comprises I image or the P image that deducts 2 values that obtain less than the group number from the group memory, and control device control apparatus for controlling of supply is to supply with decoding device with the compressed coding video frequency data of first codec processing unit and the compressed coding video frequency data of second codec processing unit respectively.
Control device can form first codec processing unit and second codec processing unit, so as the sum of I image in first codec processing unit and second codec processing unit or P image do not become than the time on formerly codec processing unit or formerly the I image of image sets or the total big predetermined number of P image or more on the time.
According to a further embodiment of the invention, a kind of information processing method is provided, is recorded in program and program on the recording medium, may further comprise the steps: obtain the playback speed instruction; Determine to be included in the decoding processing order of the image of the codec processing unit that is used for carrying out decoding processing, so as before the B image decoding I image and P image; According to the playback speed instruction that is obtained, from the image that is included in codec processing unit, select the image that to export; I image and the P image that control is selected in will the image from be included in codec processing unit and selecting as the B image supply decoder of wanting output image; The image of decoder is supplied with in decoding; And the output of control chart picture, described image is selected from decoded picture as the image that will export.
According to above-mentioned information processor, information processing method, be recorded in the program on the recording medium, a kind of program, the instruction of acquisition playback speed, determine the order of the decoding processing of image, wherein said image packets is contained in the codec processing unit that is used for carrying out decoding processing, so that decoding I image and P image before the B image, according to the playback speed instruction that is obtained, the image that selection will be exported among the image from be included in codec processing unit, according to the playback speed instruction that obtains, the image that selection will be exported among the image from be included in codec processing unit, and the image of decoding supply decoder, and output is selected as the image of wanting output image among the image that will decode.
According to a further embodiment of the invention, provide a kind of information processor, comprising: storage device is used to store the video data of compressed coding; Read-out device is used to read the video data of compressed coding; Decoding device, what be used for decoding compression-coded is video data; Apparatus for controlling of supply, be used to control the compressed coding of reading from read-out device video data be supplied to decoding device; Control device is used to control the processing of trip apparatus for controlling of supply and decoding device execution; And the playback speed command device, be used for playback speed is instructed apparatus for controlling of supply.Control device determines to be included in the order of decoding processing of the image in the codec processing unit so that carry out decoding processing by decoding device, thus before the B image decoding I image and P image.Control device is according to the playback speed instruction that is provided by the playback speed command device, and selecting from the image that is included in codec processing unit will be from the image of decoding device output.Control device control apparatus for controlling of supply is controlled I image and the P image of selecting as in the image that will comprise from the image of decoding device output during being included in codec processing unit, and the B image be supplied to decoding device.Control device control decoding device is supplied with the image of decoding device with decoding under the control of apparatus for controlling of supply, and control is as the output of the image that will select from the image of decoding device output.
According to a further embodiment of the invention, provide a kind of information processing method, be recorded in the program on the recording medium, and program, may further comprise the steps: read the video data that is stored in the compressed coding in the memory cell; The instruction of supply playback speed; The playback speed that obtains to be supplied with instructs; Determine to be included in the order that the picture decoding of the codec processing unit that is used for carrying out decoding processing is handled, so as before the B image decoding I image and P image; According to the playback speed instruction that is obtained, from the image that is included in codec processing unit, select the image that to export; Control as the image that will export and the selected I image of image from the codec processing unit that is included in the video data that reads compressed coding and P image and B image be supplied to decoder; The image of decoder is supplied with in decoding; And control is selected as the output of wanting the image of output image from the image of decoding.
According to above information processor, information processing method, with the program that is recorded on the recording medium, and program, read the video data of the compressed coding of being stored, provide and obtain the playback speed instruction, determine to be included in the order of decoding processing of the image of the codec processing unit that is used for carrying out decoding processing, so that decoding I image and P image before the B image, according to the playback speed instruction that is obtained, the image that selection will be exported from the image that is included in codec processing unit, control I image and P image, and as wanting output image and the B image from the codec processing unit that is included in the video data that reads compressed coding, selected be supplied to decoder, the image of decoder is supplied with in decoding, and output selects conduct to want the output of the image of output image from the image of decoding.
According to a further embodiment of the invention, provide a kind of information processor, comprising: decoding device is used for the video data of decoding compression-coded; Apparatus for controlling of supply, be used to control compressed coding video data be supplied to decoding device; And control device, be used to control the processing of carrying out by apparatus for controlling of supply and decoding device.Control device determines to be included in the order of decoding processing of the image of the codec processing unit that is used for carrying out the decoding processing that decoding device carries out, so as before the B image decoding I image and P image.Control device is according to the playback speed instruction that the playback speed command device provides, and selecting in being included in the image of codec processing unit will be from the image of decoding device output.Control device control apparatus for controlling of supply, be controlled in the image that is included in the codec processing unit, I image and P image and the B image selected as the image of exporting from decoding device be supplied to decoding device.Control device is under the control of apparatus for controlling of supply, and the image of decoding device is supplied with in the decoding of control decoding device, and control is as the image output of the image that will select from decoding device output.
Information processor also comprises the group memory, is used to store the image of decoding device decoding, and the group memory has than the picture number group number still less that is included in the codec processing unit.Decoding device can be carried out decoding processing as the reference image as required by the decoded picture that use is stored in the group memory.
The group memory can be stored the picture number than the number big 2 that is included in I image in the codec processing unit and P image.
The group memory can be stored 8 width of cloth images.
I image and P image can be stored in the fixed position of group memory I image and P image.
When the stem image of codec processing unit was I image from the decoded picture of decoding device output or P image, control device control decoding device was so that the demonstration output timing that the decoding that decoding device begins to decode begins regularly and decoding device begins to illustrate the image of decoding is offset first image that pre-determines number each other.
First pre-determines the total big by 1 of comparable I image of number and P image.
When the stem image of codec processing unit was B image between the decoded picture of decoding device output, control device control decoding device was offset second image that pre-determines number each other so that the demonstration that the decoding that decoding device begins to decode begins regularly and decoding device begins the image of output decoder illustrates timing.
Second pre-determines the total big by 2 of comparable I image of number and P image.
When the playback speed of playback speed indicating device indication and playback direction were normal and forward respectively, control device control decoding device was so that the decoding that decoding device begins to decode begins regularly and decoding device begins to export and exports timing through the demonstration of decoded picture and be offset first image that pre-determines number each other.
Comparable I image of the first predetermined number and P image total big by 1.
When the playback speed of playback speed indicating device indication and playback direction when being respectively normal and reverse, control device control decoding device illustrates timing and is offset second image that pre-determines number each other so that the decoding that decoding device begins to decode begins demonstration that timing and decoding device begin the image of output decoder.
Comparable I image of the second predetermined number and P image total big by 2.
Control device can be determined the timing of decoding device execution decoding processing, so that image is early wanted in the timing of the B image of the timing ratio output decoder of decoding device decoding B image.
The video data of compressed coding can comprise GOP.When reception comprises I image and P image, has than deducting 3 values that obtain by the group number from the group memory more during the image sets of big figure, control device is divided into codec processing unit with image sets, each codec processing unit comprises that the group number that has by from the group memory deducts 2 acquisition values I image and P image still less, and the control apparatus for controlling of supply is to supply with decoding device with the compressed coding video frequency data of codec processing unit.
Control device can form codec processing unit so that be included in I image in the codec processing unit of division or the sum of P image can not become than the image sets or the total big predetermined number of the I image in formerly the codec processing unit or P image or more on the time that comprise in time formerly.
Compressed coding video frequency data can comprise GOP.When reception comprises I image and P image, have than deducting 3 values that obtain by group number more during the image sets of peanut from the group memory, go up control device detection time continuously in the structure of second image sets of first image sets, if and the sum that is included in I image in first image sets and second image sets or P image is less than by deducting 3 values that obtain from the group number of group memory, control device merges first image sets and second image sets forming codec processing unit so, and the control apparatus for controlling of supply is to supply with decoding device with the video data of the compressed coding of codec processing unit.
Control device can form codec processing unit so that be included in I image in the codec processing unit of merging and the sum of P image can not become than the total big predetermined number of I image among the GOP that comprises in time formerly and P image or more.
Compressed coding video frequency data can comprise GOP.Control device can detect the structure of a GOP and the structure of the 2nd GOP in first image sets continuous in time, if and the sum that is included in I image among a GOP and the 2nd GOP or P image is less than by deducting two times of 3 values that obtain from the group number of group memory, control device merges a GOP and the 2nd GOP so, and divide the GOP merged then to form first codec processing unit and second codec processing unit, each codec processing unit comprises that the group number that is less than by from the group memory deducts the I image and the P image of 2 values that obtain, and described control device control apparatus for controlling of supply is to supply with decoding device with the video data of the compressed coding of the video data of the compressed coding of first codec processing unit and second codec processing unit respectively.
Control device can form first codec processing unit and second codec processing unit, so as the sum of I image that comprises in first codec processing unit and second codec processing unit or P image can not become than on the time formerly codec processing unit or on the time formerly the I image of image sets or the sum of P image pre-determine number or more greatly.
According to above-mentioned information processor, determine to be included in the order of decoding processing of the image of the codec processing unit that is used for carrying out the decoding processing that decoder carries out, so that decoding I image and P image before the B image, according to the playback speed that is obtained, selecting among the image in being included in codec processing unit will be from the image of decoder output, to be included among the image of codec processing unit, I image and P image, and the B image of selecting as the image that will export from decoding is supplied to decoder, the image of decoding device is supplied with in decoding, and output is as the image that will select from the image of decoding device output.
As mentioned above, according to the embodiment of the invention, can carry out quick replay operations, reverse reproduction operation or snap back replay operations to compressed coded data.Specifically, to I image and P image, and the B image of selecting as the image that will export decodes, and only exports the image that the conduct selected among the decoded image will be exported.Therefore can carry out quick replay operations, reverse reproduction operation or snap back replay operations to the video data of compressed coding, and need not to carry out complicated group memory control operation.
Description of drawings
Fig. 1 is a block diagram, illustrates the structure of replay device according to embodiment of the present invention;
Fig. 2 is a block diagram, illustrates the detailed structure of decoder shown in Figure 1;
Fig. 3 is a flow chart, illustrates control and treatment;
Fig. 4 is a flow chart, illustrates the frame of carrying out and handle 1 on each frame;
Fig. 5 is a flow chart, illustrates the inlet flow state and changes processing;
Fig. 6 is a flow chart, illustrates the decoding plan and handles 1;
Fig. 7 illustrates the long GOP of MPEG;
Fig. 8 A and 8B illustrate codec processing unit, carry out decoding processing by a decoder;
Fig. 9 A and 9B illustrate the information that is stored in the input picture formation;
Figure 10 A and 10B illustrate the information that is stored in the display sequence information formation;
Figure 11 A and 11B illustrate the information that is stored in the formation of I/P picture decoding;
Figure 12 A and 12B illustrate and are stored in DISPLAY ORDER information in the formation is set;
Figure 13 and 14 illustrates decoding timing and Displaying timer;
Figure 15 illustrates the flow chart that input is handled;
Figure 16 illustrates definite flow chart of handling of demonstration stage;
Figure 17 illustrates the flow chart that processing is determined in the temporal information plan;
Figure 18 is the flow chart that illustrates frame control and treatment 1;
Figure 19 illustrates the flow chart that I image/P picture decoding is handled;
Figure 20 illustrates the flow chart that the B picture decoding is handled;
Figure 21 and 22 illustrates decoding and Showplan;
Figure 23 illustrates to omit to handle 1 flow chart;
Figure 24 illustrates the flow chart that temporal information is reset and handled;
Figure 25 illustrates in * 2 replay operations, before the omission of carries out image and decoding afterwards and Showplan;
Figure 26 illustrates in *-2 replay operations, before the omission of carries out image and decoding afterwards and Showplan;
Figure 27 illustrates in * 4 replay operations, before the omission of carries out image and decoding afterwards and Showplan;
Figure 28 illustrates in *-4 replay operations, before the omission of carries out image and decoding afterwards and Showplan;
Figure 29 illustrates in * 5 replay operations, before the omission of carries out image and decoding afterwards and Showplan;
Figure 30 illustrates in *-5 replay operations, before the omission of carries out image and decoding afterwards and Showplan;
Figure 31 illustrates the group control in * 1 replay operations;
Figure 32 illustrates the group control in * 2 replay operations;
Figure 33 illustrates the group control in *-1 replay operations;
Figure 34 illustrates the group control in *-2 replay operations;
Figure 35 illustrates the flow chart that underflow is handled;
Figure 36 illustrates the time counter when underflow does not take place;
Figure 37 illustrates the adjustment of the time counter when underflow takes place;
Figure 38 illustrates the flow chart that a frame delay is handled;
Figure 39 illustrates frame to handle 2 flow chart;
Figure 40 is the flow chart that illustrates decoding plan 2;
Figure 41 illustrates to omit to handle 2 flow chart;
Figure 42 is the flow chart that illustrates frame control and treatment 2;
Figure 43 illustrates to omit to handle 3 flow chart;
The GOP that Figure 44 and 45 illustrates having a large amount of anchors (anchor) frame decodes;
The GOP that Figure 46 and 47 illustrates having a small amount of anchor frame decodes;
Figure 48 to 50B illustrates by dividing the decoding unit that GOP forms;
Figure 51 to 53B illustrates by merging the decoding unit that GOP forms;
Figure 54 to 59B illustrates by merging and divide the decoding unit of GOP formation;
Figure 60 illustrates the flow chart that GOP division/merging is handled;
Figure 61 and 62 illustrates by decoding unit and carries out decoding processing, and described decoding unit forms by dividing GOP;
Figure 63 and 64 illustrates by decoding unit and carries out decoding processing, and described decoding unit forms by merging and dividing GOP;
Figure 65 is the block diagram that illustrates the structure of personal computer.
Embodiment
Before describing embodiment of the present invention, below discuss in the claim feature and in embodiment of the present invention the corresponding relation between the disclosed element-specific.This description intention guarantees to have described the working of an invention mode required for protection of supporting in this specification.Therefore, even the element in the following execution mode does not relate to that certain feature of the present invention describes, this neither mean that this element does not relate to that feature of claim.On the contrary, though this relate to claim certain feature description element, that needn't mean that also this element does not relate to the further feature of claim.
Information processor (for example, replay device 1 shown in Figure 1) according to embodiment of the present invention comprising: at least one decoding device (for example, decoder 22,23 shown in Figure 1 or 24) is used for the video data of decoding compressed coding; Apparatus for controlling of supply (for example, PCI bridge 17 shown in Figure 1) is used to control the decoding device that is supplied to compression encoded video data; And control device (for example, CPU20 shown in Figure 1), be used to control the processing of apparatus for controlling of supply and decoding device execution.Control device determine the decoding processing that decoding device is carried out order in case in being included in the image of codec processing unit (for example, in the forward playback operation, comprise among the GOP except 13 width of cloth images of preceding two B images and follow hard on first three I image of the 2nd GOP of a GOP with coded sequence, 16 width of cloth images altogether of B image and B image, and in the reverse reproduction operation, first three the I image that comprises the 3rd GOP before 13 width of cloth images except that preceding two B images and the 2nd GOP among the 2nd GOP, the B image, 16 width of cloth images altogether with the B image) so that carry out decoding processing by decoding device, decoding I image and P image before the B image.The playback speed instruction that provides according to the playback speed command device, control device is selected from the image that is included in codec processing unit and comprises will be from the image of decoding device output.
Decoding device can comprise: the decoding processing final controlling element (for example, decoding processor 77 shown in Figure 2), be used to carry out decoding processing and image apparatus for controlling of supply (basic stream shown in Figure 2 address determining unit 73), what be used for the control chart picture is supplied to the decoding processing final controlling element.The image apparatus for controlling of supply can be under the control of control device, between the image in being included in codec processing unit, I image and P image and as the B image that will select from the image of decoding device output.The image that the decoding processing final controlling element can be decoded and be provided control device to supply with from image, and can under the control of control device, export as will be from the image of the image selection of decoding device output.
A plurality of decoding devices can be provided.Information processor can also comprise output switch apparatus (for example, selector 25 shown in Figure 1), is used to receive the not packed data from a plurality of decoding device outputs, and is used for optionally exporting the not packed data that is received.Control device can also be controlled the processing of being carried out by output switch apparatus.
Decoding device can comprise group memory (for example, video group memory 82 shown in Figure 2), is used to store decoded picture, and the group memory has some groups that are included in the picture number in the codec processing unit that are less than.As required, be stored in the decoded picture of group in the memory as the reference image by use, decoding device can be carried out decoding processing.
A kind of information processing method of another execution mode according to the present invention, the program and the program that are recorded on the recording medium may further comprise the steps: obtain playback speed instruction (for example, step S101 shown in Figure 15); Determine to be included in the image in the codec processing unit decoding processing order (for example, in forward playback operation, comprise among the GOP in 13 width of cloth images except that preceding two B images and the decoding order 16 width of cloth images altogether of first three I image, B image and the B image of the 2nd GOP of a GOP subsequently; In the reverse reproduction operation, 16 width of cloth images altogether that comprise first three I image of the 3rd GOP, B image and B image before 13 width of cloth images except that preceding two B images among the 2nd GOP and the 2nd GOP), so that carry out decoding processing so that before the B image, decode I image and P image (for example, step S75, S78 to S80 in Fig. 6); The image (for example, the processing shown in Figure 23,41 or 43) that image letter selection from be included in codec processing unit will be exported according to the speed command that is obtained; Control is selected among will the image from be included in codec processing unit (for example to supply with decoder as the I image that will export and P image and B image, decoding processor 77 shown in Figure 2) (for example, the step S105 among the Figure 15 after the step S196 of Figure 18); The image (for example, step S192 or S193 among Figure 18) of decoder is supplied with in decoding; And the output of the image that will export that control is selected among the image that will decode (for example, the step S196 of Figure 18 or the processing shown in the Figure 38 after the step S546 at Figure 42).
A kind of according to another embodiment of the present invention information processor (for example, replay device 1 shown in Figure 1) comprising: storage device (for example, hard disk drive shown in Figure 1 (HDD)) is so that store the video data of compressed coding; Read-out device (for example, south bridge 15 shown in Figure 1) is so that read the video data of compressed coding from storage device; Decoding device (for example, decoder 22,23 or 24 shown in Figure 1) is so that the video data of decoding compression-coded; Apparatus for controlling of supply (for example, PCI bridge 17 shown in Figure 1) is so that the video data of the compressed encoding that control will be read by read-out device is supplied with decoding device; Control device (for example, CPU 20 shown in Figure 1) is so that the processing that control is carried out by apparatus for controlling of supply and decoding device; And playback speed command device (CPU 11 for example shown in Figure 1) is so that offer control device with the playback speed instruction.Control device determine to be included in codec processing unit terminal image decoding processing order (for example, in the forward playback operation, comprise 13 width of cloth images except preceding two B images among the GOP and in coded sequence, follow first three I image of the 2nd GOP of a GOP closely, the B image, 16 width of cloth images altogether of B image, and in reverse reproduction operation, comprise 13 width of cloth images except preceding two B images among the 2nd GOP and prior to first three I image of the 3rd GOP of the 2nd GOP, the B image, 16 width of cloth images altogether of B image) so that carry out the decoding processing of carrying out by decoding device so that before the B image, decode I image and P image.The playback speed instruction image that selection will be exported from decoding device among being included in the image of codec processing unit that control device provides according to the playback speed command device.Control device control apparatus for controlling of supply is controlled among the image that is included in the codec processing unit, as the decoding device that is supplied to from the I image of decoding device output and P image and B image.Control device is under the control of apparatus for controlling of supply, and the image of decoding device is supplied with in control decoding device decoding, and can control the output as the image that will select from the image of decoding device output.
A kind of in accordance with another embodiment of the present invention information processing method, a kind of program and a kind of program that is recorded on the recording medium may further comprise the steps: read the video data (for example, at the step S1 shown in Fig. 3) that is stored in the compressed coding in the memory cell; Supply with playback speed instruction (for example, at step S5 shown in Fig. 3 or S11); The playback speed instruction (for example, at the step S101 shown in Fig. 5) that acquisition is supplied with; (for example determine to be included in order that the picture decoding of the codec processing unit that is used for carrying out decoding processing handles, in the forward playback operation, comprise 13 width of cloth images except preceding two B images among the GOP and in coded sequence, follow first three I of the 2nd GOP of a GOP closely, B, 16 width of cloth images altogether of B image, and in the reverse reproduction operation, comprise 13 width of cloth images except preceding two B images among the 2nd GOP and prior to first three I of the 3rd GOP of the 2nd GOP, B, 16 width of cloth images altogether of B image), so that at decoding I image and P image (for example, step S75 among Fig. 6 before the B image, step S78 to S80); According to the playback speed instruction that is obtained, from the image that is included in codec processing unit, select the image (for example, the processing shown in Figure 23,41 or 43) that to export; Control as the image that will export and the selected I image of image from the codec processing unit that is included in the video data that reads compressed coding and P image and B image (for example be supplied to decoder, decoding processor 77 among Fig. 2) (for example, in Figure 18 among the Figure 15 after the step S196 step S105); And the image (for example, step S192 or S193 among Figure 18) of decoder is supplied with in decoding; And control is selected from the image of decoding as the output of wanting the image of output image (for example, step S196 or the processing shown in the Figure 38 after the step S546 in Figure 42 in Figure 18).
A kind of according to another embodiment of the present invention information processor (for example, replay device 1 shown in Figure 1) comprising: decoding device (for example, decoding processor 77 shown in Figure 2) is used for the video data of decoding compression-coded; Apparatus for controlling of supply (for example, basic stream shown in Figure 2 address determining unit 73), be used to control compressed coding video data be supplied to decoding device; And control device (for example, CPU20 shown in Figure 1), be used to control the processing of carrying out by apparatus for controlling of supply and decoding device.Control device determine to be included in the codec processing unit that is used for carrying out the decoding processing that decoding device carries out image decoding processing order (for example, in the forward playback operation, comprise 13 width of cloth images except preceding two B images among the GOP and in coded sequence, follow first three I image of the 2nd GOP of a GOP closely, 16 width of cloth images altogether of B image and B image, and in the reverse reproduction operation, comprise 13 width of cloth images except preceding two B images among the 2nd GOP and prior to first three I image of the 3rd GOP of the 2nd GOP, 16 width of cloth images altogether of B image and B image), so that before the B image, decode I image and P image.Control device is according to the playback speed instruction that the playback speed command device provides, and selecting in being included in the image of codec processing unit will be from the image of decoding device output.Control device control apparatus for controlling of supply, be controlled in the image that is included in the codec processing unit, I image and P image and the B image selected as the image of exporting from decoding device be supplied to decoding device.Control device is under the control of apparatus for controlling of supply, and the image of decoding device is supplied with in the decoding of control decoding device, and control is as the image output of the image that will select from decoding device output.
Information processor can also comprise group memory (for example, video group memory 82 shown in Figure 2), is used to store the image of decoding device decoding, and the group memory has than the picture number group number still less that is included in the codec processing unit.The decoded picture that decoding device can be stored in the group memory by use is as required carried out decoding processing as the reference image.
Below by illustrating of preferred implementation, be described in detail with reference to the attached drawings the present invention.
Fig. 1 is a block diagram, for example understands the hardware configuration of replay device 1 according to embodiment of the present invention.
In replay device 1, be connected to the reading of CPU (CPU) data of 11 control stores in hard disk drive (HDD) 16 of north bridge 12, perhaps generate and output command with the beginning, the change that instruction cause CPU 20 is provided carries out with finish processing, such as control decoding plan, decoding with show.Be connected to the north bridge 12 of periphery component interconnection/interface (PCI) bus 14, under the control of CPU 11, receive the data that are stored among the HDD 16, and the data that received are supplied with memory 18 via pci bus 14 and pci bus 17 via south bridge 15.North bridge 12 also is connected to memory 13 and sends and receive the required data of processing of CPU 11.
Memory 13 is fast access memory devices, such as double data rate (DDR), wherein can store the required data of processing that CPU11 carries out.The flow data of compressed encoding has been stored in reading and writing to the data of HDD 16 of the data of south bridge 15 control from HDD 16 among the HDD 16.
PCI bridge 17 has built-in commands buffer 31 and results buffer 32, and is connected to memory 18, wherein is buffered in the flow data that reads from HDD 16 under the control of CPU 11.PCI bridge 17 can will be supplied with memory 18 from the flow data that HDD16 reads under the control of CPU 11, and with the flow data storage wherein.PCI bridge 17 also can read the flow data that is stored in the memory 18 under the control of CPU 20, and flow data is supplied with decoder 22,23 and 24.PCI bridge 17 is also via PCI bridge 14 or control bus 19, and control is corresponding to the transmission and the reception of order or result's control signal.
Commands buffer 31 also will be ordered from commands buffer 31 and read into CPU 20 via control bus 19 via the order that north bridge 12 and pci bus 14 receive from CPU 11.Results buffer 32 receives from the result of CPU 20 with via control bus 19 response commands, and the result that also will be stored in the results buffer 32 reads CPU11 via pci bus 14 and north bridge 12.
Memory 18, it for example is synchronous dynamic random access memory (SDRAM), under the control of PCI bridge 17, the flow data of the compressed code that storage is read from HDD 16.
CPU 20 reads the order that is write the commands buffer 31 of PCI bridge 17 by CPU 11 via control bus 19, and according to the processing of reading order control by PCI bridge 17, decoder 22 to 24 and selector 25 execution.The required data of processing that memory 21 storage CPU 20 carry out.
Decoder 22,23 and the 24 compressed encoding flow datas of under the control of CPU 20, decoding and being received, and output uncompressed video signal.Decoder 22,23 and 24 has internal memory 41,42 and 43 respectively, to store the flow data of the compressed encoding that is received and the vision signal of decoding as required.Decoder 22,23 and 24 needn't be included in the replay device 1.That is to say that decoder 22,23 and 24 can provide discretely with replay device 1.
Selector 25 can switch from decoder 22,23 and 24 outputs that supply with, that do not compress each frame (image) of serial digital interface (SDI) data.
Replay device 1 shown in Figure 1 can form an equipment or a plurality of equipment.For example, CPU11, north bridge 12, memory 13, south bridge 15 and HDD16 can provide the assembly as personal computer, and the function of pci bus 14, PCI bridge 17, memory 18, control bus 19, CPU20, memory 21, decoder 22,23 and 24, selector 25 can be provided in expansion card (such as pci card or PCI-express card) or expansion board.Then, this expansion card inserts in the personal computer, thereby realizes the function of replay device 1.Above-mentioned element can be included in most equipment to form replay device 1.
The operation of the replay device 1 of said structure is below described.
In HDD 16, stored the video data that utilizes the long GOP method compression of MPEG.
CPU 11 control south bridges 15 are to read the encoded data stream of compression from HDD16 via north bridge 12, to respond from the input of user from input operation unit (not shown), and supplying with the reading flow data to memory 18, and reading of data is stored in the memory 18 via north bridge 12, pci bus 14 and PCI bridge 17.CPU11 also will represent playback speed and playback direction and decoding initiation command or show that the information of initiation command writes in the commands buffer 31 of PCI bridge 17 via north bridge 12 and pci bus 14.
According to the order of supplying with and being stored in from CPU 11 commands buffer 31, CPU 20 is that the flow data of compressed encoding is determined decoding and Showplan.Or rather, the decoder 22,23 or 24 that the CPU20 selection is used to decode, and determine to be input to the timing of flow data of the compressed encoding of selected decoder 22,23 or 24, the decoding of each frame regularly, the setting of the group position that reference picture is stored, the distribution of group memory in the decoding processing, and the output of the image of decoding, i.e. Displaying timer.
CPU 20 control PCI bridges 17 flow data that will be stored in the compressed encoding in the memory 18 is supplied with selected decoder 22,23 and 24 based on determined plan then.
CPU20 control decoder 22,23 or 24 is to decode the flow data of institute's compressed and supplied coding.Decoder 22,23 or 24 decoded stream data are to generate unpressed SDI data and it is outputed to selector 25 then.
CPU 20 control selectors 25 switch the output of not compressing the SDI data from decoder 22,23 or 24 then.
Fig. 2 is a block diagram, illustrates the detailed structure of decoder 22,23 and 24.
Input processor 71 is supplied with Memory Controller 74 with the flow data of compressed code from PCI bridge 17, and flow data is stored in the input buffer 75.Input processor 71 also obtains header addresses, data size, image header information, picture size information and the Q matrix of each image from the flow data of supplying with, and with the information providing address administration table 72 that is obtained.
The above-mentioned information stores of each image that address administration table 72 will be supplied with from input processor 71 be the table information that can be identified by Table I D.
Basic stream address determining unit 73 reads the header addresses and the picture size information of respective image from table information, described table information is represented by the predetermined Table I D that is stored in the address administration table 72, and according to the control signal of supplying with from CPU 20 via control bus 19 with the header addresses and the picture size information providing Memory Controller 74 that are read, so that the flow data that is stored in the input buffer 75 can be supplied with decoding processor 77 with elementary area.
Memory Controller 74 control is from/reading and writing to the flow data of input buffer 75.That is to say, Memory Controller 74 will write input buffer 75 from the flow data that input processor 71 is supplied with, also, predetermined image is read into decoding processor 77 according to the header addresses and the picture size information of the correspondence image of supplying with from basic stream address determining unit 73.
Input buffer 75 is corresponding to the partial record zone of memory 41,42 shown in Figure 1 and 43, and described memory is SDRAM, and under the control of Memory Controller 74 the storage flow data.
Control bus 76 will be supplied with basic stream address determining unit 73, decoding processor 77, write image address determining unit 78, reference picture address determining unit 79 and OPADD determining unit 80 via the control signal that control bus 19 receives from CPU 20, and the information that also will be referred to the processing of execution in decoder 22,23 or 24 is supplied with CPU 20 via control bus 19.
As required by the corresponding reference image of reference from 79 supplies of reference picture address determining unit, the mpeg video stream that decoding processor 77 is decoded and read from input buffer 75 under the control of Memory Controller 74, and base band (compression) vision signal of decoding supplied with write image address determining unit 78.
Write image address determining unit 78 and obtain the control signal via control bus 19 supplies from CPU 20 via control bus 76, and determine record position according to this control signal, that is, in video group memory 82 from the storage sets position of decoding processor 77 decoding and the baseband video signal supplied with.Write image address determining unit 78 then via the group location storage baseband video signal determined of Memory Controller 81 at video group memory 82.
Reference picture address determining unit 79 obtains via the control signal of control bus 19 from the CPU20 supply via control bus 76, and according to this control signal control storage controller 81, with from video group memory 82, read be stored in the group of reference picture formerly of being appointed as the P image group in frame image data or be stored in be appointed as frame image data in the group of reference picture group formerly and subsequently of B image, and the frame image data that is read is supplied with decoding processor 77.
OPADD determining unit 80 obtains the control signal via control bus 19 supplies from CPU20 via control bus 76, and specify the group of input picture according to the frame image data of this control signal from be stored in video group memory 82, promptly, the group that is used for the frame that will show, and control storage controller 81 is read input picture.
Memory Controller 81 control from/read and write to the two field picture of video group memory 82.Video group memory 82 is corresponding to the partial record zone of memory 41,42 shown in Figure 1 and 43, described memory is SDRAM, and video group memory 82 is 8 framing image sets memories, is used for the group of describing before under the control of Memory Controller 81 each frame of view data being stored in.
The control and treatment of carrying out below with reference to the flow chart description CPU 11 of Fig. 3.
In step S1, CPU 11 control north bridges 12 and south bridge 15 come to read from HDD 16 a plurality of GOP of the compressed encoding flow data that will decode and export of user's appointment.
In step S2, CPU 11 supplies with PCI bridge 17 via pci bus 14 with the reading flow data, and flow data is sent to memory 18.
In step S3, by via north bridge 12 and pci bus 14 data being transmitted the commands buffer 31 of finishing message and image information supply PCI bridge 17, data transmission end and image information that CPU 11 will be referred to be included in the image among the GOP that sends memory 18 to send to CPU 20.Image information comprises for example image type, the header message that relates to each image header and picture size.
In step S4, CPU 11 is from CPU 20 and memory 18 ready for receiving message.Or rather, CPU 11 reads the result via pci bus 14 and north bridge 12 and finishes message and image information with response from the data transmission that CPU 20 supplies with results buffer 32 via control bus 19, also receives the message that expression GOP flow datas have been stored from memory 18 via PCI bridge 17, pci bus 14 and north bridge 12.
In step S5, CPU 11 receives instruction to begin the output processing of resetting via operation input unit (not shown) from the user, and send the decoding initiation commands to commands buffer 31, and cause decoder 22,23 or 24 beginning decoding processing via north bridge 12 and pci bus 14.The decoding initiation command comprises display speed information.
In step S6, CPU 11 sends via north bridge 12 and pci bus 14 and shows that initiation command is to commands buffer 31.In step S7, CPU 11 begins to show a GOP by the sdi signal of corresponding decoder decode, that is, and and the baseband images signal.
In step S8, CPU 11 detects and has shown a GOP.Or rather, CPU 11 examines then by read the result via pci bus 14 and north bridge 12 which image has shown, supply with the demonstration initiation command of results buffer 32 via control bus 19 from CPU 20 with response, be that message is finished in the demonstration that writes results buffer 32 by CPU 20 among the step S194 in Figure 18 discussed below, it represents that each frame all showed.
In step S9, CPU 11 determines whether shown GOP is last GOP of flow data.If finding shown GOP is last GOP, processing finishes so.
If in step S9, determine, shown GOP is not last GOP, handle so and advance to step S10, determine whether to change inlet flow state (for example finishing the playback stream data) from operation input unit (not shown) signal supplied with basis by user input instruction, change the flow data of resetting, perhaps change playback speed or direction.
If in step S10, find to change the instruction of inlet flow state, handle so and advance to step S11, wherein CPU 11 sends ordering to commands buffer 31 corresponding to user's operation via north bridge 12 and pci bus 14.
If find also that at step S10 not input changes the instruction of inlet flow state, perhaps after step S11, handle so and advance to step S12 to determine in HDD16, whether having any flow data that will show.If in step S12, determine, there is not the flow data that will show, handle turning back to step S7 sends to memory 18 with processing flow data so.
If determine in step S12, the flow data that existence will show is handled so and is advanced to step S13, and wherein CPU 11 control north bridges 12 and south bridge 15 are read the GOP subsequently of flow data from HDD16.
In step S14, CPU 11 supplies with PCI bridge 17 via pci bus 14 with the GOP that reads, and GOP is sent to memory 18.That is to say, in memory 18, except that when having stored the afterbody of the flow data that will reset, stored the GOP that pre-determines number basically.
In step S15, finish message and image information is supplied with commands buffer 31 by via north bridge 12 and pci bus 14 data being transmitted, CPU 11 sends and relates to the data that are included in the image among the GOP that will send memory 18 to and transmit and finish message and image information to CPU 20.Image information comprises for example type and the size of each image.
In step S16, CPU 11 is from CPU 20 and memory 18 ready for receiving message.Or rather, CPU 11 reads the result via pci bus 14 and north bridge 12, finish message and image information with response from the data transmission that CPU20 supplies with results buffer 32 via control bus 19, CPU 11 also receives expression GOP flow data from the message of memory 18 via PCI bridge 17, pci bus 14 and north bridge 12 storages.
After step S16, handle advancing to step S7, and carry out with reprocessing.
According to above-mentioned processing, order to CPU 20 by supply, and reception result is with the order supplied with of response the demonstration of the data of the Code And Decode that CPU 11 can the control flows data.
In this is handled, in the GOP unit, show the data of decoding.When output by the sdi signal of corresponding decoder decode during to external source, CPU 11 can carry out the processing that is similar to above-mentioned processing.Therefore, supply with the order that CPU 20 and reception result are supplied with response by ordering, the data of the Code And Decode that CPU 11 can the control flows data are to the output of external source.
According to the order of supplying with from CPU 11, CPU 20 controls are by the decoding processing of a plurality of decoders ( decoder 22,23 and 24 among Fig. 1) execution then.Or rather, CPU 20 selects the decoder 22,23 or 24 that is used to decode, promptly supplies with the decoder 22,23 or 24 of the video data of compressed encoding.CPU 20 also according to flow data to the incoming timing of selected decoder, for the decoding of each image regularly, the setting of reference picture position in the group, in decoding processing the distribution of group memory and decoded picture output regularly (being Displaying timer) control PCI bridge 17, decoder 22,23 and 24 and selector 25.The decoding of being carried out by CPU 20 has been discussed and has been shown the output control operation below with reference to Fig. 4 to Figure 38.
Memory 21 has a plurality of registers and message queue so that store various types of information, to allow CPU 20 control decoders 22,23 and 24 decoding processing of carrying out and to handle from the selection of the sdi signal of selector 25 outputs.Message queue is first in first out (FIFO) formation, wherein regularly plan or decoding/demonstration control purpose according to decoding, can store the order supplied with from CPU 11, the image I D relevant and have the information (for example, temporal information) of image I D with the predetermined degree of depth with image type information.
CPU 20 controls decoding and Displaying timer by the information that use is stored in the corresponding informance formation.That is to say, substitute the formation image in the memory 21, be stored in the corresponding informance formation, calculate so that control decoding and Displaying timer thereby allow CPU 20 to carry out at the control hypograph of CPU 20.
Storage various types of information and the message queue that is used for CPU 20 various control operations comprise: command queue, the order that storage obtains from commands buffer 31 via control bus 19; The input picture formation is wherein with the image I D of coded sequence storage corresponding to the input picture of formation; The image I D in the input picture formation of being stored in DISPLAY ORDER rearrangement has wherein been stored in the display sequence information formation; I image and P image that the image I D from be stored in the display sequence information formation extracts have wherein been stored in the formation of I/P picture decoding, and rearrange with decoding order; The formation of temporal information I/P picture decoding wherein the image I D in being arranged on the formation of I/P picture decoding, has been stored the temporal information corresponding to image I D; DISPLAY ORDER is provided with formation, wherein with DISPLAY ORDER image I D is set; The temporal information DISPLAY ORDER is provided with formation, wherein is provided with the image I D in the formation except that being arranged on DISPLAY ORDER, has also stored the temporal information relevant with image I D; And the demonstration formation, the image I D of the image that storage will show subsequently.The information that is stored in each formation, the degree of depth of formation and the details of passing through the processing of use information and executing below are discussed.
Handle 1 below with reference to the frame that the flow chart description of Fig. 4 is carried out on each frame.For each frame repeats this processing routine until the treated flow data that will show, perhaps until the instruction of finishing display operation is provided.
In step S31, carry out the inlet flow state of describing in detail below with reference to Fig. 5 and change processing.Change in the processing at the inlet flow state, CPU 20 checks the newer command of sending from CPU 11.
In step S32, carry out the decoding plan and handle 1, describe this processing in detail below with reference to Fig. 6.Handle in 1 in the decoding plan, arrange decoding plan regularly.
In step S33, the time counter in the processing time of CPU 20 referential expressions one frame determines, one of first to the 6th frame of the flow data whether the theme frame will be reset with decoding order.
If determine that in step S33 the theme frame is not one of first to the 6th frame of flow data, handle so and advance to step S34, carry out a frame delay in this step and show set handling, below with reference to Figure 38 this processing is described.In a frame delay shows set handling, the base band two field picture that is arranged among the step S35 of first pre-treatment routine decoding in the frame control and treatment 1 and generates.
If determine that in step S33 the theme frame is one of first to the 6th frame of flow data, perhaps after step S34, handle and advance to step S35, wherein carry out frame control and treatment 1, will this processing be discussed with reference to Figure 18.In frame control and treatment 1, the frame to be processed of decoding.
In step S36, CPU 20 determines that all frames have all handled.If determine that at step S36 not every frame has all been handled, handle so and advancing to step S37, wherein CPU 20 increases progressively time counter with frame unit.Then, handle and turn back to step S31, repeat processing subsequently then.
If determine that in step S36 all frames are all treated, processing finishes so.
According to processing shown in Figure 4, CPU 20 increases progressively time counter for each frame, and arranges the plan of decoding processing according to the playback speed of user indication, and allows corresponding decoder each frame of decoding.
The inlet flow state of discussing below with reference to the flow chart of Fig. 5 among the step S31 of Fig. 4 changes the details of handling.
In step S51, CPU 20 checks the command queue of storage from the order of commands buffer 31.
In step S52, CPU 20 judges whether the newer command that changes inlet flow state such as change playback speed and playback direction or finish replay operations is stored in the command queue.
If in step S52, determine, stored newer command, handle advancing to step S53 so, wherein CPU 20 changes the inlet flow state according to the order that earliest time is stored in the command queue, for example playback speed and playback direction.If in step S52, determine in command queue, not store new order or in step S53 after, handle turning back to step S31 among Fig. 4 so.
Handle according to this, check that earliest time is stored in the order in the command queue, the order that the storage of described command queue obtains from commands buffer 31, and, change the inlet flow state according to this order.
1 details is handled in the decoding plan of carrying out among flow chart description Fig. 4 step S32 below with reference to Fig. 6.
In step S71, CPU 20 determines with reference to the input picture formation whether the input picture formation is empty.The input picture formation is message queue, and the input that described message queue is carried out in step S72 is provided with in handling, the image I D (will be the plan of decoding of described input picture) and the required image of scheme of arrangement of its storage input picture.
Supposition now, the flow data of handling in replay device 1 is the long GOP flow data of MPEG, each GOP comprises 15 width of cloth images, and is as shown in Figure 7 all.Each decoder 22,23 and 24 of carrying out the decoding processing respectively organize 15 width of cloth images receives 13 width of cloth images and first three width of cloth IBB image (under the situation of reverse reproduction operation) 16 width of cloth images altogether of first three width of cloth IBB image of GOP (under the situation of forward playback operation) or previous GOP subsequently of preceding two width of cloth B images that remove DISPLAY ORDER of a GOP.
Below with reference to Fig. 8 A and 8B the decoding unit of distributing to decoder 22,23 and 24 is discussed.In Fig. 8 A and 8B, show the arrangement of the image of supplying with decoder 22,23 and 24 with DISPLAY ORDER.In replay device 1, each GOP with 15 width of cloth images decodes in a decoder 22,23 and 24.As mentioned above, each decoder 22,23 and 24 receives 13 width of cloth images and first three width of cloth image 16 width of cloth images altogether of GOP before or subsequently of preceding two width of cloth B images that remove DISPLAY ORDER of a GOP.That is to say, if the playback direction is a forward, shown in Fig. 8 A, so first decoders of selecting from decoder 22,23 and 24 receive a GOP except 13 width of cloth images of preceding two width of cloth B images and first three width of cloth image of the 2nd GOP, and second decoder receive the 2nd GOP except 13 width of cloth images of preceding two width of cloth B images of DISPLAY ORDER and first three width of cloth image of the 3rd GOP.Similarly, the 3rd decoder receive the 3rd GOP except 13 width of cloth images of preceding two width of cloth images and first three width of cloth image of the 4th GOP.If it is reverse playing direction, shown in Fig. 8 B, first decoders of selecting from decoder 22,23 and 24 receive the 2nd GOP except 13 width of cloth images of preceding two width of cloth B images and first three width of cloth image of the 3rd GOP, and second decoder receive a GOP except 13 width of cloth images of preceding two width of cloth B images and first three width of cloth image of the 2nd GOP.
In the input picture formation, be provided with and for the image I D of the input picture of its plan of decoding and be used for the required image of scheme of arrangement.That is to say, in the input picture formation, stored the image I D of 18 width of cloth images altogether, be i.e. 15 width of cloth images of a GOP of the long GOP flow data of MEPG and before or subsequently the image I D of first three width of cloth IBB image of GOP.
Fig. 9 A for example understands the image I D of first three width of cloth IBB image of 15 width of cloth images of the GOP be stored in the long GOP flow data of MPEG in the input picture formation and the 2nd GOP following closely, operates with the execution forward playback.Fig. 9 B for example understand the 2nd GOP be stored in the long GOP flow data of MPEG in the input picture formation 15 width of cloth images and prior to the 2nd GOP the image I D of first three width of cloth IBB image of the 3rd GOP, operate to carry out reverse reproduction.
If determine at step S71, the input picture formation is not empty, handles the step S32 that turns back to Fig. 4 so.That is to say that in decoding plan was handled, 16 width of cloth images altogether promptly, from 13 width of cloth images of the 3rd width of cloth image to the 15 width of cloth images of GOP and first three width of cloth image of GOP subsequently, were handled as processing unit.
If determine at step S71, the input picture formation is empty, carries out input so and handle in step S72, below with reference to the flow chart of Figure 15 described processing is discussed.
In step S73, CPU 20 determines according to the register value of the decoder of expression reception subsequent data whether the temporal information order of delay is provided with formation is empty, and it is to be provided with according to the decoder that receives subsequent data that described temporal information order of delay is provided with formation.It is to determine to handle the formation that is provided with in the temporal information plan that is used for a plurality of decoders (Fig. 1 decoder 22,23 and 24) of step S80 that the temporal information DISPLAY ORDER is provided with formation.Flow chart below with reference to Figure 17 provides the details that processing is determined in the temporal information plan.
If in step S73, determine, temporal information DISPLAY ORDER corresponding to the decoder that receives subsequent data is provided with formation not for empty, that is to say,, handle the step S32 that turns back to Fig. 4 so if handle over against every two field picture of the GOP of scheme of arrangement is carried out decoding processing or shown.
If determine in step S73, the temporal information DISPLAY ORDER is provided with formation for empty, handles so and advances to step S74, and wherein CPU 20 carries out the processing of reordering.In the processing of reordering, arrange the image I D that is arranged on 18 width of cloth images in the input picture formation with coded sequence and rearrange, and be arranged in the display sequence information formation with DISPLAY ORDER.
Therefore, if the playback direction is a forward, be set in the display sequence information formation as the image I D that is arranged in the input picture formation that is discussed with reference to figure 9A, rearrange so with the DISPLAY ORDER shown in Figure 10 A.If the playback direction is oppositely, be set in the display sequence information formation as the image I D that is arranged in the input picture formation that is discussed with reference to figure 9B, rearrange so with the DISPLAY ORDER shown in Figure 10 B.
Handle when in the display sequence information formation, image I D being set when carrying out at step S74 to reorder, all images ID that output is arranged in the input picture formation, and the input picture formation becomes sky.That is to say that if determine, the input picture formation is empty, means the image I D that has resequenced and arranged in the input picture formation so in step S71.If in step S71, determine, the input picture formation is not empty, mean that so the input processing 1 that is provided with among the step S72 of formation in the plan of execution early decoding is handled owing to the temporal information DISPLAY ORDER corresponding to the decoder that receives subsequent data is not sky afterwards, thereby do not have the processing of reordering among the execution in step S74.
In step S75, the display sequence information formation that CPU 20 references are provided with in step S74, and be arranged on the I image of the GOP that will decode with decoding order in the formation of I/P picture decoding and the image I D of P image.The formation of I/P picture decoding is such formation, wherein is provided with prior to the I image that will decode of B image and the image I D of P image with decoding order.
Figure 11 A and 11B for example understand the I/P picture decoding formation when carrying out quick forward playback operation and reverse reproduction operation respectively.In the formation of I/P picture decoding, when carrying out the forward playback operation, shown in Figure 11 A, the image I D corresponding to 6 width of cloth images of the I image of the image I D that is provided with and P image is set in display sequence information formation shown in Figure 10 A.In the formation of I/P picture decoding, when carrying out the reverse reproduction operation, shown in Figure 11 B, the image I D corresponding to 6 width of cloth images of the I image of the image I D that is provided with and P image is set in display sequence information formation shown in Figure 10 B.
In step S76, the reference picture group of the P image that is used to decode is also specified in the group position of CPU 20 assigned I images and P image storage.For I image and P image use 6 fixing groups of 8 groups in the video group memory 82 shown in Figure 2.
By at first decode I image and P image (being also referred to as " anchor frame ") and by the I image of fixing decoding and the group position of P image storage, can be from any image replaying flow data in the forward playback operation, in the reverse reproduction operation or in the random playback operation.By 6 width of cloth I images and the P image of at first decoding, because I image and P treatment of picture time are shorter, thereby from the user provide the beginning replay operations or change playback speed or the time of the instruction of playback direction to can reducing, and no matter the position of demonstration start frame among the GOP by the time of reflecting this instruction display image.In addition, can reduce the demonstration time of wiping demonstration desired image in the replay operations.As a result, can improve performance during the quick replay operations.
In step S77, the position of CPU 20 designated reference image sets so as according to as the B image of decoding in the group position of specified storage I image of step S76 and P image.The group position of specifying the B image to be stored in frame control and treatment 1 below is discussed.
In step S78, CPU20 is provided with DISPLAY ORDER formation is set.
Figure 12 A and 12B understand for example that respectively the DISPLAY ORDER in forward playback operation and reverse reproduction operation neutralization is provided with formation.Be provided with in the formation at DISPLAY ORDER, when carrying out the forward playback operation, shown in Figure 12 A, be arranged on 18 width of cloth image I D that arrange in the display sequence information formation as shown in Figure 10 A with DISPLAY ORDER, except preceding two width of cloth B images and last I image, just, 15 width of cloth image I D, promptly, supply with 16 width of cloth image I D of one of decoder 22,23 and 24 shown in Fig. 8 A, except last I image.
Be provided with in the formation at DISPLAY ORDER, when carrying out the reverse reproduction operation, shown in Figure 12 B, be arranged on 18 width of cloth image I D that arrange in the display sequence information formation as shown in Figure 10 B with DISPLAY ORDER, except preceding two width of cloth B images and last I image, just, 15 width of cloth image I D, promptly, supply with 16 width of cloth image I D of one of decoder 22,23 and 24 shown in Fig. 8 B, except last I image.
When in step S78, DISPLAY ORDER being set formation is set, the image I D that arranges in the display sequence information formation is all exported, and the display sequence information formation becomes sky.
In step S79, carry out the demonstration stage of discussing and determine to handle below with reference to Figure 16.Determine in the processing that in the demonstration stage determining to decode in the processing unit begins timing and show the skew (displacement) that begins between the timing, it equals the GOP and 1 frame that comprise 15 frames.
In step S80, carry out the temporal information plan of discussing below with reference to the flow chart of Figure 17 and determine to handle.In the temporal information plan determine to be handled, for the decoding of controlling each frame regularly, the plan of decoding, and in the predetermined message queue relevant, temporal information is set with image I D.
In step S79 and S80, arrange the plan of decoding timing and Displaying timer.Or rather, when the stem of display image is I image or P image, mean that replay operations direction and playback speed are forward and normal.In this case, decoding I image and P image before the B image, as shown in figure 13, and plan decoding regularly and Displaying timer so as to show beginning regularly and decoding begin timing slip 6 width of cloth images, and cause the Displaying timer of B image and frame of decoding timing slip of B image.When the stem of display image is the B image, mean that the playback direction is that oppositely in other words, speed is provided with the expression negative value.In this case, I image and P image were decoded before the B image, as shown in figure 14, and arrange the plan of decoding timing and Displaying timer, and cause the Displaying timer of B image and frame of decoding timing slip of B image so that the demonstration beginning regularly begins timing slip 7 width of cloth images with decoding.
In step S81, CPU 20 switches the decoder that receives subsequent data.Or rather, when the register value that receives the decoder of subsequent data when expression equaled the value (that is, being 2 in the replay device 1 shown in Figure 1) of certain ratio decoder device number (shown in Figure 1 in replay device 1 3) little 1, it was 0 that CPU 20 is provided with register value.When register value ratio decoder device number little 2 or more for a long time, CPU 20 increases progressively 1 with register value.After step S81, handle turning back to step S32 shown in Figure 4.
According to above-mentioned processing, arrange the plan of decoding timing and Displaying timer.
With reference to the flow chart of Figure 15, provide the description that performed input is handled among the step S72 among Fig. 6.
In step S101, CPU20 obtains to be included among the step S5 of Fig. 3 by CPU11 and sends to display speed information in the decoding initiation command of commands buffer 31, perhaps import, send to order acquisition speed value of setting of commands buffer 31, and in memory 21, store the information that is obtained by CPU 11 Fig. 3 step S11 response user.If the speed value of setting is for just, the direction of resetting so is a forward, and if speed be set to negative, the direction of resetting so is for oppositely.
Following definite speed value of setting.If the speed value of setting is 1, carry out the normal playback operation so.If the speed value of setting greater than 1, is carried out quick replay operations so.If the speed value of setting, is carried out the low-speed reproducing operation so less than 1.If the speed value of setting is-1, carry out the reverse reproduction operation of normal speed so.If the speed value of setting less than-1, is carried out the snap back replay operations so.If the speed value of setting is an absolute value less than 1 negative value, carry out the low-speed reverse replay operations so.
Then, be plus or minus according to the speed value of setting that in step S101, obtains, CPU 11 determines that in step S102 the processing that will carry out is forward playback operation or reverse reproduction operation.
If find that in step S102 the playback direction is a forward, handle advancing to step S103 so.In step S103, CPU 20 in the input picture formation of discussing with reference to figure 9A, be provided with the GOP that will decode from 13 frames of stem and the image I D of I image, B image and the B image of GOP subsequently.
If determine that in step S102 the playback direction is not a forward, the direction of promptly resetting is oppositely, handles advancing to step S104.In step S104, CPU 20 is provided with the image I D from I image, B image and the B image of 13 frames of afterbody and previous GOP of the GOP that will decode in the input picture formation of discussing with reference to figure 9B.
After step S103 or S104, in step S105, receive the register value of the register of subsequent data according to the setting of input picture formation and expression, the view data that CPU 20 control PCI bridges 17 compress with 16 frames in the flow data that will be stored in memory 18, that discuss with reference to figure 8A and 8B is sent to the predetermined decoder of selection from decoder 22,23 and 24.In this case, for a GOP of flow data, register value is the value of determining in initial setting up, and for second or GOP subsequently, register value is the register value of determining in the step S81 of Fig. 6.
In decoder 22,23 or 24, input processor 71 is supplied with Memory Controller 74 with 16 frame data, and with storage in input buffer 75.Input processor 71 also will be referred to header addresses, data size, the picture size information of each image for example, the information providing address administration table 72 of Q matrix, and with the information stores of each image table information for identifying by Table I D.
In step S106, be sent to the results buffer 32 of the result of predetermined decoder by representing a GOP stream via control bus 19 supply PCI bridges 17, CPU 20 sends the stream transmission and finishes message to CPU 11.Handle the step S72 that turns back to Fig. 6 then.
Handle according to input, 16 frame data that serve as codec processing unit as with reference to figure 8A and 8B discussion offer decoder 22,23 or 24.
Be provided with respect to decoding beginning constant time lag below with reference to the flow chart description of Figure 16 being used to of in the step S79 of Fig. 6, carrying out and show that the demonstration stage in beginning stage regularly determines to handle.
In step S131, CPU 20 stores the information that the picture number that will show that is provided with in the formation is set about DISPLAY ORDER shown in Figure 12 A and the 12B in display image number register.Because the DISPLAY ORDER in Figure 12 A and 12B is provided with the image I D that is provided with 15 width of cloth images in the formation, thereby is arranged in the display image number register 15.
In step S132, CPU 20 will be stored in the I image that will decode in the formation of I/P picture decoding shown in Figure 11 A and the 11B and the sum of P image is stored in the I/P picture number register.Owing in the I/P picture decoding formation shown in Figure 11 A and the 11B, be provided with the image I D of 6 width of cloth images, thereby be stored in the I/P picture number register with 6.
In step S133, CPU 20 determines, being arranged on the image type that DISPLAY ORDER is provided with the image I D of formation stem is the B image.In other words, determine that when calculating the normal speed stage skew in the replay operations forward or backwards, the playback direction is a forward or oppositely at step S133.Handling after 1 carrying out the following omission that will describe, is the B image even be arranged on the image type that DISPLAY ORDER is provided with the image I D of formation stem, and according to playback speed, the playback direction can be a forward or reverse.
If discovery is the B image at the image type that DISPLAY ORDER is provided with the image I D of formation stem in step S133, handles so and advancing to step S134.In step S134, the sum that CPU 20 is provided with by the I image that will decode and P image increases by 1 number that obtains, begins regularly, at DISPLAY ORDER the time phase offset d isp_phase that the demonstration of the image of the image I D that is provided with in the formation begins timing is set as the decoding with respect to the I image of the image I D that is provided with in the formation of I/P picture decoding and P image.
For example, when playback speed be *-1 the time, as shown in figure 14, the image type of image I D that the stem of formation is set at DISPLAY ORDER is the B image.After decoding I image and P image, decoding B image is so that decoding timing slip 7 width of cloth images of the Displaying timer of B image and I image and P image, and causes the Displaying timer of B image and frame of decoding timing slip of B image.
If determine that at step S133 the image type that the image I D of formation stem is set at DISPLAY ORDER is not the B image, handle so and advance to step S135.At step S135, the I image that CPU 20 will decode and the sum of P image are set to that decoding with respect to the I image of the image I D that is provided with and P image begins regularly, the demonstration that the image of the image I D that is provided with in the formation is set at DISPLAY ORDER begins time phase offset d isp_phase regularly in the formation of I/P picture decoding.
For example, when playback speed be * 1 the time, as shown in figure 13, the image type of image I D that the stem of formation is set at DISPLAY ORDER is not B image but I image.After decoding I image and P image, decoding B image is so that decoding timing slip 6 width of cloth images of the Displaying timer of I image and I image and I image and P image, and causes the Displaying timer of B image and frame of decoding timing slip of B image.
After step S134 or S135, in step S136, the demonstration stage offset d isp_phase that CPU 20 will determine in step S134 or S135 is stored in the built-in register.
In step S137, CPU 20 determines whether theme GOP shows initial GOP (comprising that wherein theme GOP is located at the situation that the playback speed change restarts the position of replay operations afterwards).
If determine in step S137, theme GOP shows initial GOP, handles advancing to step S138 so.In step S138, CPU 20 deducts stage skew prev_disp_phase among the previous GOP in formerly handling from the demonstration stage offset d isp_phase that determines among step S134 or S135, and end value is stored in the built-in register as offset correction values disp_zero.The demonstration stage offset d isp_phase substitution that CPU 20 also will determine in step S134 or S135 is the skew of the demonstration stage among GOP prev_disp_phase formerly.Handle then and turn back to step S79 among Fig. 6.
If determine in step S137, theme GOP shows initial GOP, handles advancing to step S139 so, and wherein CPU 20 is arranged on initial value 0 in the built-in register as offset correction values disp_zero.Handle the step S79 that turns back to Fig. 6 then.
Determine to handle according to the demonstration stage, can determine to show the beginning stage regularly with respect to decoding beginning constant time lag.
Describing the temporal information plan of carrying out among the step S80 of Fig. 6 below with reference to the flow chart of Figure 17 determines to handle.
In step S161, CPU 20 determines whether theme GOP changes the inlet flow state codec processing unit corresponding to a GOP afterwards.The codec processing unit that its image type has changed after carries out image is omitted is not first codec processing unit but second or codec processing unit subsequently.Therefore, if determine that in step S161 theme GOP is the codec processing unit corresponding to a GOP, handle so and advance to step S166.
If determine that in step S161 theme GOP is not the codec processing unit corresponding to a GOP, handle so and advance to step S162 to determine whether first image has been the B image from I image or P image modification after the omission of carries out image.
If determine in step S162, first image has been the B image from I image or P image modification after carries out image is omitted, and handles advancing to step S163 so, and wherein CPU 20 stage adjusted value is set to 1.Processing advances to step S167.
If in step S162, determine, first image is not the B image from I image or P image modification also after carries out image is omitted, and handles so to advance to step S164 to determine that after the carries out image omission whether first image is I image or P image from the B image modification.
If determine in step S164, first image is handled so and is advanced to step S165 through being I image or P image from the B image modification after carries out image is omitted, and wherein CPU 20 stage adjusted value is set to-1.Handle then and advance to step S167.
If determine that in step S161 theme GOP is first codec processing unit, if determine in step S164 that perhaps first image is not I image or P image from the B image modification also after carries out image is omitted, handle so and advance to step S166, wherein CPU 20 stage adjusted value is set to 0.
After step S163, S165 or S166, handle advancing to step S167.At step S167, shown in following equation (1), by decode start time information prev_time_base, the stage offset correction values disp_zero that uses the early decoding processing unit, the temporal information adjusted value added_count that in omission processing 1 (will be described below), calculates and the stage adjusted value of in the step of step S161 to S166, determining, CPU 20 calculates decode start time information time_base, and it represents the time that first frame (this is a codec processing unit) of 16 width of cloth images is decoded in one of decoder 22,23 and 24:
The decode start time information prev_time_base-stage offset correction values disp_zero+ temporal information adjusted value added_count+ stage adjusted value of the previous GOP of decode start time information time_base=... (1)
In order to calculate the decode start time information that shows initial GOP by use equation (1), the decode start time information prev_time_base of previous GOP and the temporal information adjusted value added_count that calculates in the omission processing 1 that is used for previous GOP are set to 0.
In step S168, CPU 20 is provided with the formation of temporal information I/P picture decoding for the decoder that receives follow-up data.
In the formation of temporal information I/P picture decoding, the decode start time information time_base of the image that will at first decode and being used for counts the count value of the time counter of every frame and is arranged to be associated with the image I D that is arranged on the I/P picture decoding formation shown in Figure 11 A and the 11B.For each decoder 22,23 and 24 provides the formation of temporal information I/P picture decoding.
Or rather, receive the register value of the decoder of subsequent data by referential expression, CPU 20 is provided for receiving the temporal information I/P picture decoding formation of the decoder of subsequent data.CPU 20 uses decode start time information time_base to be arranged in the information of time of frame of the first image I D of the image I D in the formation of I/P picture decoding corresponding to decoding as expression, and service time counter count value as the information of expression decoding corresponding to the time of the frame of other image I D.
When the formation of temporal information I/P picture decoding is set in step S168, owing to exported all images ID that is arranged in the formation of I/P picture decoding, thereby the formation of I/P picture decoding becomes sky.
In step S169, CPU 20 is provided with formation for the decoder that is used to receive subsequent data is provided with the temporal information DISPLAY ORDER.
Be provided with in the formation at the temporal information DISPLAY ORDER, begin regularly to deduct 1 and the count value of the value that obtains and time counter is set to that the image I D that is provided with the formation is set is relevant with the DISPLAY ORDER of being discussed with reference to figure 12A and 12B with respect to decoding beginning stage skew regularly from the demonstration of first image that will show.For each decoder 22,23 and 24 of replay device 1 provides the temporal information DISPLAY ORDER that formation is set.
Or rather, receive the register value of the decoder of subsequent data by referential expression, the temporal information DISPLAY ORDER that CPU 20 is provided for receiving the decoder of subsequent data is provided with formation.CPU 20 uses by deducting 1 value that obtains as adjusting to B picture decoding reference time information regularly from stage offset d isp_phase, described stage offset d isp_phase is illustrated in the Displaying timer of first frame in the DISPLAY ORDER that the demonstration stage determines in determine handling, and service time counter count value as the information of expression decoding corresponding to the time of the frame of other image I D.
When the temporal information DISPLAY ORDER being set in step S169 formation is set, be arranged in DISPLAY ORDER all images ID in the formation is set owing to exported, thereby DISPLAY ORDER is provided with formation and becomes sky.
In step S170, CPU 20 will be added to the decode start time information prev_time_base that decode start time information time_base goes up the previous GOP of value conduct that obtains by a picture number (it is stored in the display image number register) that will show in the step S131 of Figure 16, be stored in the register.Handle the step S80 that turns back to Fig. 6 then.
Determine to handle according to the temporal information plan, the decoding processing timing of 16 frames (for codec processing unit) is set.
According to determining to handle referring to figs. 16 and 17 the demonstration stage of being discussed respectively and the definite processing of temporal information plan, based on the number of I image and P image, the picture number that will show and first image that will show is the B image, and decode phase and demonstration stage are set to temporal information.Therefore, dynamically change speed, can constantly show decoded picture, above-mentioned change takes place when using a plurality of decoders to carry out quick forward and reverse reproduction operation by change according to the picture number that in each decoder, will decode and show.Even changed the omission cycle, also can in the corresponding speed scope, constantly change playback speed by increasing or reducing and want displayed map as number for each frame in response to beginning to change the instruction of playback speed from certain image.
According to reference to figure 6 to 17 processing of being discussed, the decoding plan in can the step S32 of execution graph 4 is handled.
Then, if determine that in step S33 the theme frame is one of first to the 6th frame, handle so by the demonstration of the frame delay among skips steps S34 set handling and advance to step S35.In step S35, carry out frame control and treatment 1.
Below with reference to the flow chart of Figure 18 the frame control and treatment 1 that step S35 carries out among Fig. 4 is discussed.
In step S191, by reference time information I/P picture decoding formation be used for managing the time management counter of the timing of each processing that replay device 1 carries out, CPU 20 determines whether to exist over and done with any flow data of demonstration time.If in step S191, determine, there is over and done with flow data of demonstration time, handle advancing to step S197 so.
If in step S191, determine, there is not over and done with flow data of demonstration time, handle advancing to step S192 so, wherein carry out I image/P picture decoding of discussing below with reference to Figure 19 and handle.
In step S193, carry out the B picture decoding of discussing below with reference to Figure 20 and handle.
In step S194, by displays image information is write the results buffer 32 of PCI bridge 17 via control bus 19, CPU 20 sends the demonstration initiation command that displays image information send from CPU 11 with step S6 in the response diagram 3 to CPU 11.This makes CPU 11 understand which image that will show which GOP by the displays image information that reference is stored in the results buffer 32.
In step S195, CPU 20 increases progressively the time management counter.
Then, in step S196, carry out the omission of describing below with reference to Figure 23 and handle 1, and handle the step S35 that turns back to Fig. 4 then.
If in step S191, determine, there is its demonstration time flow data of process, handle so and advance to step S197, wherein carry out the underflow of describing below with reference to Figure 35 and handle.Handle the step S36 that turns back to Fig. 4 then.
According to the frame control and treatment, if there is no its demonstration time flow data of process so based on frame of decoding plan decoding, and is carried out to omit and is handled 1.If there is its demonstration time flow data of process, carries out underflow described below so and handle.
With reference now to the flow chart of Figure 19, provides the description of I image/P picture decoding processing of carrying out among the step S192 among Figure 18.
In step S221, it is relevant and be arranged on temporal information in the formation of temporal information I/P picture decoding that CPU 20 determines whether the value of time management counters is matched with image I D with the image that will decode subsequently.If determine that in step S221 the value of time management counter does not match temporal information, handle so and turn back to step S192 among Figure 18.
If determine that in step S221 the value of time management counter is matched with temporal information, handle so and advance to step S222.In step S222, CPU 20 is via control bus 19 control decoders 22,23 or 24 (being appointed as the decoder of carrying out decoding), with decoding I image or P image, and from the formation of temporal information I/P picture decoding the image I D of deletion decoded picture.
Or rather, CPU 20 referential expressions receive the register value of the decoder of subsequent data, and the basic stream address determining unit 73 of control respective decoder, allowing Memory Controller 74 from input buffer 75, to read view data, and reading of data is supplied with decoding processor 77 corresponding to the image I D that in the formation of temporal information I/P picture decoding, is provided with.
If the image of decoding is the I image, CPU 20 control decoding processors 77 are decoded from the I image of Memory Controller 74 supplies so, also control is write image address determining unit 78 frame data of decoding is supplied with Memory Controller 81, and the I image of decoding is stored among Fig. 6 step S76 is appointed as in the group of I image storage sets.If the image of decoding is the P image, so CPU 20 control reference picture address determining units 79 with according to the P image reference group position that in Fig. 6, is provided with among the step S76, under the control of Memory Controller 81, read the reference picture that is stored in the video group memory 82, and the reference picture that reads is supplied with decoding processor 77, and 77 decodings of control decoding processor are from the P image of Memory Controller 74 supplies.CPU 20 also controls and writes image address determining unit 78 the frame data of decoding are supplied with Memory Controller 81 and the P image of decoding is stored in the group of being appointed as P image storage sets that is provided with among the step S76 of Fig. 6.
After step S222, handle the step S192 that turns back to Figure 18.
Handle according to I image/P picture decoding, can be based on the plan of regulation decode I image and P image.
With reference now to the flow chart of Figure 20, the description that the B picture decoding of carrying out among the step S193 in Figure 18 that provides is handled.
In step S241, CPU 20 determines whether the value of time counter is consistent with the temporal information that the first image I D that is provided with in the formation is set corresponding to the temporal information DISPLAY ORDER.Being arranged on the temporal information DISPLAY ORDER, temporal information in the formation is set is to subtract 1 value that obtains (adjusting to the decoding reference time information regularly of B image) by the expression determined determining to handle from the demonstration stage with the stage offset d isp_phase of the Displaying timer of first frame in 15 frames of DISPLAY ORDER, perhaps this temporal information be be different from DISPLAY ORDER in the count value of the relevant time counter of the image I D of frame of first frame.If determine in step S241, the value of time counter is handled so and is turned back to step S193 among Figure 18 not with consistent corresponding to the temporal information that the first image I D that is provided with in the formation is set at the temporal information DISPLAY ORDER.
If determine at step S241, the value of time counter is with at the temporal information DISPLAY ORDER temporal information that is provided with in the formation to be set consistent, handling so and advance to step S242, is the B image with the value uniform images ID that determines its time information and time management counter.
If in step S242, determine, corresponding image I D is the B image, handle so and advance to step S243, wherein CPU 20 is via control bus 19 control decoders 22,23 or 24 (it is appointed as the decoder of carrying out decoding), with the B image of decoding and being represented by correspondence image ID.
Or rather, CPU 20 is with reference to the value that is stored in the register, expression receives the decoder of subsequent data, and the basic stream address determining unit 73 of controlling corresponding decoder reads in the view data that the temporal information DISPLAY ORDER is provided with the image I D that is provided with in the formation to allow Memory Controller 74 from input buffer 75, and supplies with decoding processor 77 with the view data that will read.CPU 20 controls reference picture address determining units 79 and reads the reference picture that is stored in the video group memory 82 to allow Memory Controller 81 according to the reference group position of the B image that is provided with among the step S77 in Fig. 6 then, and the reference picture that reads is supplied with decoding processor 77.CPU 20 control decoding processors 77 are with the B image of decoding from Memory Controller 74 supplies then.
In step S244, CPU20 is provided with the group position of B image storage.Or rather, CPU20 control is write image address determining unit 78 so that in video group memory 82, and the B image alternately is stored in two groups not being appointed as I image/P image storage sets.CPU 20 controls are write image address determining unit 78 with the frame data supply Memory Controller 81 with decoding processor 77 decodings, and the frame data of decoding are stored in the group position that is provided with in the video group memory 82.
If determine in step S242, corresponding image I D is not the B image, perhaps after step S244, handles advancing to step S245.In step S245, CPU 20 will be arranged on the temporal information DISPLAY ORDER and the first image I D in the formation is set is arranged on and shows in the formation.Show that formation is the formation with the degree of depth of only storing an image I D.
Or rather, if determine in step S242, corresponding image I D is not the B image, and in step S245, the I image that CPU 20 will decode in the step S222 of Figure 19 or the image I D of P image are arranged on and show in the formation so.If determine in step S242, corresponding image I D is the B image, handles advancing to step S243, the B image of wherein decoding so.In this case in step S245, CPU 20 will be arranged on the temporal information DISPLAY ORDER and the first image I D (that is the image I D of the B image of decoding) in the formation is set is arranged on and shows in the formation.
From the temporal information DISPLAY ORDER output being set the formation is arranged on the temporal information DISPLAY ORDER the first image I D in the formation is set, and the image I D of image that will follow the image of the image I D that is provided with closely in showing formation is arranged on the stem that the temporal information DISPLAY ORDER is provided with formation, and perhaps the temporal information DISPLAY ORDER is provided with formation and becomes sky.After step S245, handle turning back to step S193 among Figure 18.
As mentioned above, carry out decoding according to being arranged on the temporal information that temporal information I/P picture decoding formation and temporal information DISPLAY ORDER be provided with in the formation.In forward playback operation, just as shown in Figure 21, according to being arranged on temporal information in the formation of temporal information I/P picture decoding decode I image and P image, and temporal information in the formation B image of decoding is set according to being arranged on the temporal information DISPLAY ORDER.In reverse reproduction operation, as shown in figure 22, according to being arranged on temporal information in the formation of temporal information I/P picture decoding decode I image and P image, and temporal information in the formation B image of decoding is set according to being arranged on the temporal information DISPLAY ORDER.
Shown in Figure 21 and 22, show set handling by carrying out a frame delay discussed below, the decoding that the B image is set regularly in case the decoding of B image regularly with frame of Displaying timer skew of B image, and the direction of no matter resetting is a forward or oppositely.
Figure 21 and 22 for example understand when playback speed be respectively * 1 and * decoding timing and Displaying timer-1 time.In quick replay operations, handle 1 by omission as described below and change the setting that the temporal information DISPLAY ORDER is provided with formation, and after omitting processing 1, the next execution in the reprocessing routine that be provided with that the change of formation is set according to the temporal information DISPLAY ORDER decoded.In quick replay operations, the B image of not decoding all, but the I image and the P image of decoding all.
Flow chart below with reference to Figure 23 is described in the omission processing of carrying out among the step S196 of Figure 18 1.
In step S271, CPU 20 determines that according to the inlet flow state that is provided with in the inlet flow state change processing that is arranged on reference to figure 5 playback speed that the user is provided with is quick replay operations.If determine in step S271, playback speed is not fast, handles turning back to the step S196 of Figure 18 so, and advances to step S36 among Fig. 4.
If in step S271, determine to have carried out quick replay operations, handle so and advance to step S272, to determine that in change speed back theme frame being set is first frame of a GOP.
If find that in step S272 the theme frame is first frame of a GOP, handle advancing to step S273 so, wherein CPU 20 determines the omission cycle according to the speed value of setting, and the determined omission cycle is arranged in the register.
Or rather, if playback speed be * 2 or *-2, the 20 omission cycles of CPU are set to 1/2 so, and it is stored in the register.If playback speed is * 3 or *-3, the 20 omission cycles of CPU are set to 1/3 so, and it is stored in the register.
In step S274, CPU 20 resets and is used to count the frame counter of frame number, and no matter omitting the codec processing unit of handling in 1.
If determine at step S272, the theme frame is not first frame that changes an inlet flow state GOP afterwards, or after step S274, handle advancing to step S275 and each image I D in the formation is set be provided with check mark at the temporal information DISPLAY ORDER so, be arranged on the temporal information DISPLAY ORDER image I D in the formation is set thereby determined whether complete inspection to determine whether association store.Check mark is with being arranged on the temporal information DISPLAY ORDER among the step S280 that discusses mark in the formation to be set hereinafter.If in step S275, determined on inspection to be stored in the temporal information DISPLAY ORDER all image I D in the formation are set, handle the step S196 that turns back among Figure 18 so, and advance to step S36 among Fig. 4.
If in step S275, determine, do not have review time information DISPLAY ORDER that image I D all in the formation is set, in other words, exist some that the image I D of check mark is not set, handle advancing to step S276 so, wherein CPU 20 increases progressively 1 with frame counter.
Then in step S277, CPU 20 is with reference to determining and being stored in omission cycle and frame counter value in the register, to determine whether the being presented at frame of representing in the frame counter.Or rather, if frame counter is represented 2n (n is a positive integer) when the omission cycle is 1/2, CPU20 determines to show topics frame so, and in other cases, CPU 20 determines not show topics frame.If frame counter is represented 3n (n is a positive integer) when the omission cycle is 1/3, CPU 20 determines to show topics frame so, and in other cases, CPU 20 determines not show topics frame.If in step S277, determine to show topics frame, that is, do not omit this frame, handle advancing to step S280 so.
If determine that in step S277 the frame that do not show topics promptly, omits this frame, handle so and advance to step S278.In step S278, CPU 20 is provided with the formation image I D of deletion theme frame from the temporal information DISPLAY ORDER, perhaps in the temporal information DISPLAY ORDER is provided with image I D in the formation mark (not show tags) that expression does not show this frame is set.
In step S279, carry out the temporal information replacement of describing below with reference to Figure 24 and handle.
If in step S277, determine to be presented at the frame of representing in the frame counter, perhaps after step S279, handle so and advance to step S280.In step S280, CPU 20 is provided with among the image I D that in the formation check mark is arranged on the frame of representing in the frame counter at the temporal information DISPLAY ORDER.Processing turns back to step S275, and repeats corresponding processing.
According to omit handling 1, from the temporal information DISPLAY ORDER image I D of the frame that deletion does not show the formation is set, perhaps the general not show tags be arranged among the image I D that the temporal information DISPLAY ORDER is provided with the correspondence in the formation.By reference time information DISPLAY ORDER formation is set and carries out B image demonstration processing and frame delay demonstration set handling (following discussion).Because in response to quick replay operations, upgraded the setting that the temporal information DISPLAY ORDER is provided with formation, thereby after this handles routine, only in other words the decoding B image that will show, will not supply with decoding processor 77 by abridged B image.Although, do not show them to wanting abridged I image and P image to decode.
Be independent of GOP ground and constantly the frame counter value increased progressively, and need not to reset, until having changed the speed setting, that is, and until having changed the inlet flow state.Because omitting the denominator in cycle is the value that can't divide exactly the number of the frame that forms a GOP, for example be when a GOP comprises 15 frames and playback speed * 2, * 4, *-2 or *-4 the time, the position of the image that does not show changes according to GOP sometimes.Even in this case, be illustrated in the frame counter value of the frame sequential in the inlet flow by use, also can determine whether the frame that shows topics easily according to the omission cycle.
Like this, with its omission, and omit I image and P image before with B image input decoder, and need not after with its decoding, to show, thereby realize quick replay operations by the dynamic change of the speed of following with regular intervals.Because I image and P image can be used as the reference picture of other image during decoding, thereby this method is being effective by carrying out in the quick replay operations with regular intervals omission image with the group of minimal amount in memory.
Handle 1 according to omitting, when omitting image,, so can speed be set to from *-3 to * 3 the scope if use two decoders with regular intervals, if and used three decoders, so could speed would be set to from *-6 to * 6 the scope.Be similar to above-mentioned omission and handle 1 omission and handle by using two or more decoders, can carrying out, thereby realize quick replay operations.
Handle below with reference to the temporal information replacement that the flow chart discussion of Figure 24 is carried out in the step S279 of Figure 23.
In step S301, CPU 20 detects because the change in the speed value of setting is carried out and omits the number of handling 1 that increased or the image deleted.
In step S302, the picture number that CPU 20 will increase or delete is set to temporal information adjusted value added_count.
In step S303, CPU 20 replacement temporal information DISPLAY ORDERs are provided with the temporal information in the formation, so that can constantly arrange temporal information.Handle the step S279 that turns back among Figure 23 then.
Reset to handle according to temporal information, change according to playback speed and carry out the temporal information DISPLAY ORDER that omits after handling setting in the formation is set.In with the reprocessing routine, according to the temporal information DISPLAY ORDER that changes formation is set and controls the decoding timing of B image and the Displaying timer of all frames.
Be described in the temporal information DISPLAY ORDER of omit handling in 1 below with reference to Figure 25 to 30 setting in the formation is set.The top of Figure 25 to 30 understands that for example carrying out omission handles plan situation before, the abridged frame to dot.The bottom of Figure 25 to 30 understands that for example carrying out omission handles plan situation afterwards.In Figure 25 to 30, numeral 0,1 in the bracket or 2 expressions: the information in the message queue of being stored in is perhaps carried out the information that is stored in the message queue by which decoder 22,23 or 24 corresponding to which decoder 22,23 or 24.
For example, in * 2 quick replay operations, as shown in figure 25, the change number that omits the image that will show after the image in decoding unit is-7, and therefore temporal information adjusted value added_count becomes-7.Then, because the image type of first image is the B image from the I image modification in decoding unit subsequently, thereby the stage adjusted value becomes-1.Therefore, if stage offset correction values disp_zero is 0, the amount that time_base adjusted is exactly-8 so.Because first image of the 2nd GOP is the B image, thus the demonstration of the 2nd GOP just begin regularly decoding than I image begun timing late 7 frames.
For example, in *-2 quick replay operations, as shown in figure 26, the change number that omits the image that will show after the image in decoding unit is-7, and therefore temporal information adjusted value added_count becomes-7.Then, because the image type of first image keeps identically with the B image in decoding unit subsequently, thereby the stage adjusted value becomes 0.Therefore, if stage offset correction values disp_zero is 0, the amount that time_base adjusted is exactly-7 so.Because first image of the 2nd GOP is the B image, thus the demonstration of the 2nd GOP just begin regularly decoding than I image begun timing late 7 frames.
For example, in * 4 quick replay operations, as shown in figure 27, the change number that omits the image that will show after the image in decoding unit is-11, and therefore temporal information adjusted value added_count becomes-11.Then, because the image type of first image is the B image from the I image modification in decoding unit subsequently, thereby the stage adjusted value becomes-1.Therefore, if stage offset correction values disp_zero is 0, the amount that time_base adjusted is exactly-12 so.Because first image of the 2nd GOP is the B image, thus the demonstration of the 2nd GOP just begin regularly decoding than I image begun timing late 7 frames.
For example, in *-4 quick replay operations, as shown in figure 28, the change number that omits the image that will show after the image in decoding unit is-11, and therefore temporal information adjusted value added_count becomes-11.Then, because the image type of first image keeps identically with the B image in decoding unit subsequently, thereby the stage adjusted value becomes 0.Therefore, if stage offset correction values disp_zero is 0, the amount that time_base adjusted is exactly-11 so.Because first image of the 2nd GOP is the B image, thus the demonstration of the 2nd GOP just begin regularly decoding than I image begun timing late 7 frames.
For example, in * 5 quick replay operations, as shown in figure 29, the change number that omits the image that will show after the image in decoding unit is-12, and therefore temporal information adjusted value added_count becomes-12.Then, because the image type of first image keeps identically with the I image in decoding unit subsequently, thereby the stage adjusted value becomes 0.Therefore, if stage offset correction values disp_zero is 0, the amount that time_base adjusted is exactly-12 so.Because first image of the 2nd GOP is the I image, thus the demonstration of the 2nd GOP just begin regularly decoding than I image begun timing late 6 frames.
For example, in *-5 quick replay operations, as shown in figure 30, the change number that omits the image that will show after the image in decoding unit is-12, and therefore temporal information adjusted value added_count becomes-12.Then, because the image type of first image keeps identically with the B image in decoding unit subsequently, thereby the stage adjusted value becomes 0.Therefore, if stage offset correction values dip_zero is 0, the amount that time_base adjusted is exactly-12 so.Because first image of the 2nd GOP is the B image, thus the demonstration of the 2nd GOP just begin regularly decoding than I image begun timing late 7 frames.
Then, in the step S167 that the temporal information plan of the Figure 17 that will repeat is subsequently determined to handle, shown in Figure 25 to 30, calculate decode start time information time_base according to equation (1).Then, the decoding of calculating corresponding to the I image of the image I D that is provided with in the formation of temporal information I/P picture decoding and P image according to the decode start time information time_base that is calculated begins regularly, and according to decode start time information time_base that is calculated and temporal information DISPLAY ORDER formation (having deleted the frame that does not show from this formation) is set then, determines to carry out decoding and the Showplan that the B image after 1 is handled in omission.
For example, present hypothesis, when N is 15 (N is the number of a frame among the GOP) and M when being 3 (M is the interval between I image, P image or I image and the P image), the long GOP of replay device 1 decoding MPEG flows.In this case, in 8 groups of video group memory 82 shown in Figure 2, anchor frame I, P, P, P, P, I are not stored among the group No.0 to No.5 with not changing, and the B image alternately is stored among group No.6 and the No.7, so that a Zao frame is decoded to them than showing.Utilize this configuration,, can in the velocity interval that is provided with according to decoder number (3 decoders are arranged in the replay device 1 of this execution mode), (forward and reverse) carry out display operation although the number of the group in the video group memory 82 is 8.
The setting of video group memory 82 is described below with reference to Figure 31 to 34.
Figure 31 illustrates the image in 8 groups that are stored in the video group memory 82 shown in Figure 2 when carrying out * 1 replay operations.
In the group No.0 to No.5 that is occupied by I image and P image, the I image and the P image of decoding are stored continuously with decoding order.When showing an I2 image, the B3 image is stored among the group No.6 simultaneously, and when showing the B3 image, the B4 image is stored among the group No.7 simultaneously.When demonstration B3 image, discharge the group No.6 that stores the B3 image that has shown and store the group No.0 that is used as the I2 image of reference picture.Thereafter, the B image alternately is stored among group No.6 and the No.7, so that the Displaying timer of B image is than the decoding of a B image late frame regularly, and has shown B image release group No.6 and No.7 afterwards.Release group No.1 to No.5 after the P image of being stored is used as reference picture.That is to say, after showing the image of being stored, do not discharge group No.0 to No.5, but be used as reference picture and shown these images release group No.0 to No.5 afterwards at the image of being stored.
Figure 32 illustrates the image of storing in 8 groups of video group memory 82 shown in Figure 2 when carrying out * 2 replay operations.
In the group No.0 to No.5 that is occupied by I image and P image, the I image and the P image of decoding are stored continuously with decoding order.When showing an I2 image, the B4 image is stored among the group No.6 simultaneously, and when showing the B4 image, the B6 image is stored among the group No.7 simultaneously.Then, when demonstration B6 image, discharge the group No.6 that stores the B4 image that has shown and store the group No.0 that is used as the I2 image of reference picture.Thereafter, B image in abridged alternately is not stored among group No.6 and the No.7, so that the Displaying timer of B image is than the decoding of a B image late frame regularly, and has shown B image release group No.6 and No.7 afterwards.After showing the image of being stored, do not discharge group No.0 to No.5, but be used as reference picture and shown these images release group No.0 to No.5 afterwards at the image of being stored.Therefore at the P image of being stored and I image as after the reference picture and shown these images release group No.1 to No.5 afterwards.
Figure 33 illustrates the image of storing in 8 groups of video group memory 82 shown in Figure 2 when execution *-1 replay operations.
In the group No.0 to No.5 that is occupied by I image and P image, the I image and the P image of decoding are stored continuously with decoding order.After in the I2 image being stored in group No.5, the B1 image is stored among the group No.6, and when showing the B1 image, the B0 image is stored among the group No.7 simultaneously.When showing the B1 image, discharge the group No.6 that stores the B1 image that has shown.Thereafter, the B image alternately is stored among group No.6 and the No.7, so that the Displaying timer of B image is than the decoding of a B image late frame regularly, and has shown B image release group No.6 and No.7 afterwards.After having shown the image of being stored and using the image of being stored as release group No.0 to No.5 after the reference image.That is to say that owing to carry out the reverse reproduction operation in example shown in Figure 33, last demonstration at first is stored in the I2 image among the group No.0, just release group No.0 after the demonstration of this GOP finishes.
Figure 34 illustrates when execution *-2 replay operations, is stored in the image in 8 groups in the video group memory 82 shown in Figure 2.
In the group No.0 to No.5 that is occupied by I image and P image, the I image and the P image of decoding are stored continuously with decoding order.After the I2 image being stored in group No.5, the B1 image is stored among the group No.6, and after having shown the B1 image, the B12 image is stored among the group No.7.Then, discharge the group No.6 that stores the B1 image that has shown.Thereafter, B image in abridged alternately is not stored among group No.6 and the No.7, so that the Displaying timer of B image is than the decoding of a B image late frame regularly, and has shown B image release group No.6 and No.7 afterwards.After having shown the image of being stored and using the image of being stored as release group No.0 to No.5 after the reference image.That is to say that owing to carry out the reverse reproduction operation in example shown in Figure 34, last demonstration at first is stored in the I2 image among the group No.0, just release group No.0 after the demonstration of this GOP finishes.
As mentioned above, when the view data such as the compression of the bidirectional interframe predictive of MPEG that use is generally used for method for compressing image is carried out the reverse reproduction operation or fast during replay operations, the plan of decoding is so that the Displaying timer of at first decode I image and P image and B image becomes than the decoding of a B image late frame regularly, and thereby I image and P image can be stored in fixed position in the group.Thereby, can utilize the group of minimum number to carry out reverse reproduction operation or snap back replay operations effectively, and not need the control operation of complicated group.
Or rather, be included in the I image of codec processing unit and the number and 2 group sums that are used to store the B image of P image if be used for the group number of storage frame, so just enough.In this case, by I image and P image (anchor frame) are stored in the fixed position in the group, can realize quick replay operations or reverse reproduction operation and need not the control operation of complicated group.
For example, in the long GOP stream of MPEG, when N is that 15 (N is the number of the frame in a GOP) and M are when being 3 (M is the time interval between I image, P image or I image and the P image), as this embodiment, preparation has 8 groups video group memory altogether, that is, 5 groups are used for I, P, P, P and P image, one group is used for I image subsequently, and two groups are used for the B image.Utilize this structure,, can realize quick replay operations, reverse reproduction operation or snap back replay operations, and need not the control operation of complicated group by I image and P image being stored in the fixed position in the group.
In order to improve to response from the instruction of user's change speed, the decoding of B image or hang up so that change speed when showing with frame unit.In this case, although the decoding of B image timing and Displaying timer are separate managements, the control of execution group simply.For example, even in reference to other image prediction image mistake is taking place when inlet flow changes, it recovers correct frame (underflow that describes below is handled).Thereby speed can change by frame unit.
Handle below with reference to the underflow that the flow chart description of Figure 35 is carried out in the step S197 of Figure 18.
In step S331, by reference time management counter and time counter, CPU 20 determines whether the supply of flow data postpones, for example, whether underflow has taken place with respect to showing to handle.If in step S331, determine, underflow does not take place, handle among the step S197 that turns back to Figure 18 so and advance among the step S36 of Fig. 4.
If in step S331, determine to have taken place underflow, handle so and advancing to step S332, wherein CPU20 adjusts to time in the stem of temporal information I/P image queue with the value of time counter.Handle then and turn back to step S197 among Figure 18, and advance to step S36 among Fig. 4.
The adjustment of time counter when underflow takes place is described below with reference to Figure 36 and 37.In Figure 36 and 37, numeral 0,1 in bracket or 2 expressions are stored in information in the message queue corresponding to which decoder 22,23 or 24, and the information that perhaps is stored in the message queue by which decoder 22,23 or 24 is carried out.
Figure 36 illustrate the information that is stored in the message queue and when as Figure 30 performed *-5 replay operations in underflow when not taking place, about the decoding of time counter regularly and Displaying timer.In Figure 36, by way of example, the decoding of the I image of GOP begins time counter regularly and represents 7.
For example, if the amount of two frames has taken place underflow, time counter postpones two frames so, as shown in figure 37, and by coming reboot process with reference to the time counter of adjusting, thereby allow to plan forthwith subsequently, decode and show control.
According to this processing, even the supply of stream is owing to for example postpone when the throughput of execution minimizing of HDD16 during quick replay operations, can detect the underflow that causes owing to the supply delay of flowing, and, handle thereby be not subjected to interruptedly to restart to reset with the time that the count delay underflow of temporal information takes place.
If determine among the step S36 that in Fig. 4 not every frame has all been handled, handle so and advance to step S37, wherein increase progressively time counter.Handle then and turn back to step S31, and repeat corresponding the processing.
For the 7th or subsequently frame,, show set handling so in step S34, carry out a frame delay owing to determine that in step S33 the theme frame is not one of first to the 6th frame.
With reference to the flow chart of Figure 38, be given in the description that a frame delay of carrying out among the step S34 among Fig. 4 shows set handling.
In step S361, CPU 20 sends idsplay order to respective decoder 22,23 or 24 according to being stored in the information that shows in the formation via control bus 19, so that show that handling the ratio decoder that becomes handles a late frame, and from show formation, delete corresponding image I D.In this case, respective decoder 22,23 or 24 OPADD determining unit 80 receive control signal via control bus 76 from CPU20, and control storage controller 80 is read corresponding image and the image that reads is supplied with selector 25 from video group memory 82.
In step S362, CPU20 controls the frame of selector 25 output decoders according to the register value of the decoder of expression reception subsequent data, and described register value is provided with among the step S81 in Fig. 6.Handle then and turn back to step S34 among Fig. 4.
For example, if playback speed be * 1, as shown in figure 21, following execution shows control.With the temporal information DISPLAY ORDER order display image of image in the formation is set, so that according at the temporal information DISPLAY ORDER temporal information that is provided with in the formation being set, it is regularly more late than the decoding of B image that the Displaying timer of B image becomes.
If playback speed is *-1, and as shown in figure 22, so following execution shows control.With the temporal information DISPLAY ORDER image sequence display image that is provided with in the formation is set, so that according at the temporal information DISPLAY ORDER temporal information that is provided with in the formation being set, it is regularly more late than the decoding of B image that the Displaying timer of B image becomes.
This is applied to shown in Figure 25 * 2 replay operations equally, replay operations shown in Figure 26 with *-2 speed, operating shown in Figure 27 with * 4 speed playbacks, shown in Figure 28 with *-4 speed playback operation, shown in Figure 29 with the replay operations shown in the replay operations of * 5 speed and Figure 30 with *-5 speed.
According to this processing, if the group number of storage frame equals to be included in I image and the number of P image and two group sums that are used to store the B image in the codec processing unit, by I image and P image (it is the anchor frame) are stored in fixed position in the group, can realize quick replay operations, reverse reproduction operation or snap back replay operations, and need not the control operation of complicated group.
Decoding regularly and the skew of the stage between the Displaying timer be to determine according to being included in the I image in the codec processing unit and the number of P image.As selection, the stage skew between decoding timing and the Displaying timer can be determined according to the number of the frame that will show among the frame that is included in the codec processing unit.
When carrying out fast forward or backwards replay operations, be used for second and subsequently codec processing unit decoding in the works, before the frame of frame of determining to show and not demonstration, promptly before carrying out the frame omission, can detect coding parameter in advance, for example the image type that will show in the stem of codec processing unit.
For example, before frame of determining to show and the frame that do not show, detect the image type that will show in the codec processing unit stem in advance, for example, according to the formerly frame counter of counting in the works of codec processing unit and the omission cycle of calculating.
Utilize this structure, can more accurately plan to handle with more promptly carrying out to decode.
The particular instance be used to detect in the processing of the image type of the image of wanting the demonstration of codec processing unit stem is below described.
With reference to the flow chart of Figure 39, provide the frame that will on each frame, carry out and handled 2 description, wherein can before executive plan, detect the image type of the image that the stem at codec processing unit shows.For each frame repeats this processing routine until the treated flow data that will show or until the instruction of finishing display operation is provided.
In step S431, carry out the inlet flow data mode of being discussed with reference to figure 5 and change processing.
In step S432, carry out the decoding plan of describing below with reference to Figure 40 and handle 2.Handle in 2 in the decoding plan, planned decoding regularly.Before decoding is regularly planned to handle, detect the image type of the image that shows in the codec processing unit stem.
Then, in step S433, CPU 20 is by the time counter in the processing time of each frame of referential expression, determines the frame in first to the 6th frame of the flow data whether the theme frame will be reset with decoding order.
If determine that in step S433 the theme frame is not the frame in first to the 6th frame, handle so and advance to step S434, wherein carry out a frame delay of discussing with reference to Figure 38 and show set handling.
If determine in step S433, the theme frame is the frame in first to the 6th frame of the flow data that will reset, perhaps after step S434, handles and advances to step S435, wherein carries out the frame control and treatment of discussing below with reference to Figure 42 2.In frame control and treatment 2, decoding theme frame.
CPU 20 has determined whether all frames treated in step S436.If determine that in step S436 not every frame has all passed through processing, handle so and advance to step S437, wherein CPU 20 increases progressively time counter for each frame.
After step S437, handle and turn back to step S431, and repeat processing subsequently.If determine that in step S436 all frames have all passed through processing, end process so.
As mentioned above, CPU 20 increases progressively time counter for each frame, carries out the decoding plan according to the playback speed of user's indication, and each frame of decoding.In this is handled, before carrying out decoding timing plan, detect the image type of the image that shows in the codec processing unit stem.
Be described in the decoding plan of carrying out among the step S432 among Figure 39 below with reference to flow chart among Figure 40 and handle 2.
Step S471 to S478 is the step that is similar to the step S71 to S78 in Fig. 6 basically respectively.
In step S471, CPU 20 determines by the reference input rank whether input rank is empty.If input rank is not empty, handle turning back to step S432 among Figure 39 so.If it is empty finding input rank in step S471, handle advancing to step S472 so, wherein carry out the input of discussing with reference to Figure 15 and handle.
Determine in step S473 then whether for the temporal information DISPLAY ORDER of the decoder setting that receives subsequent data formation is set is empty.If the temporal information DISPLAY ORDER is provided with formation for empty, that is to say, handle if each frame of the GOP of plan carried out decoding processing or showing, handle turning back to step S432 among Figure 39 so.
If discovery time information DISPLAY ORDER is provided with formation for empty in step S473, handle so and advance to step S474, wherein carry out the processing of reordering.
In step S475, the I image of the GOP that in the formation of I/P picture decoding, will decode and the image I D of P image with the decoding order setting.Then, in step S476, the decode reference picture group of P image of the group position of designated store I image and P image and being used to.In step S477,, specify the position of the reference picture group of the B image that is used to decode according to the group position that is used to store I image and P image.Then, be arranged on the DISPLAY ORDER shown in Figure 12 A and the 12B formation is set.
In step S479, carry out the omission of describing below with reference to Figure 41 and handle 2.Omit to handle in 2, before carrying out decoding regularly plan is handled, detecting the image type of the image that will show in the codec processing unit stem.
Step S480 to S482 is similar to the step S79 to S81 among Fig. 6 basically respectively.That is to say, in step S480, carry out the demonstration stage shown in Figure 16 and determine to handle.In step S481, carry out temporal information plan shown in Figure 17 and determine to handle timing and Displaying timer so that plan is decoded.
Or rather, because the omission among the step S479 handles 2, the image type that decoding regularly detect the image that will show in the codec processing unit stem before the plan can carried out.Therefore, for example, if the stem that shows is I image or P image, as shown in figure 13, plan decoding timing and Displaying timer are so that decoded I image and P image before decoding B image so, ratio decoder is regularly wanted late 6 width of cloth images so that Displaying timer becomes, and consequently the Displaying timer of B image becomes than the late frame of decoding timing of B image.If show that stem is the B image, as shown in figure 14, plan decoding timing and Displaying timer are so that decoded I image and P image before decoding B image so, ratio decoder is regularly wanted late 7 width of cloth images so that Displaying timer becomes, and consequently the Displaying timer of B image is than the late frame of decoding timing of B image.
In step S482, switch the setting of the decoder that receives subsequent data.Then, processing turns back to step S432 among Figure 39.
Plan decoding timing and Displaying timer according to decoding plan 2.
Handle 2 below with reference to the omission that the flow chart discussion among Figure 41 is carried out in the step S479 of Figure 40.
In step S501, CPU 20 is according to determining at the inlet flow state that changes setting in the processing with reference to figure 5 described inlet flow states whether the playback speed of user's indication is quick.If determining playback speed in step S501 is not fast, handle advancing to step S510 discussed below so.
If finding playback speed in step S501 is fast, handle advancing to step S502 so, to determine whether the theme frame is first frame of a GOP after having changed the speed setting.
If find that in step S502 the theme frame is first frame of a GOP, handle advancing to step S503 so, wherein CPU 20 determines the omission cycle according to the speed value of setting GOP_Speed and the determined omission cycle is stored in the register.
Or rather, if playback speed be * 2 or *-2, the 20 omission cycles of CPU are set to 1/2 and it is stored in the register so.If playback speed is * 3 or *-3, the 20 omission cycles of CPU are set to 1/3 and it is stored in the register so.
In step S504, CPU 20 resets and is used to count the frame counter of frame number, and no matter omitting the codec processing unit of handling in 2.
If determine in step S502, the theme frame is not first frame at a GOP after having changed the inlet flow state, perhaps after step S504, handles advancing to step S505.In step S505, by determining whether to be provided with the relevant check mark of image I D in the formation is set with being stored in DISPLAY ORDER, determined whether on inspection to be stored in DISPLAY ORDER all images ID that stores in the formation has been set.Check mark is will at DISPLAY ORDER the mark that is provided with in the formation to be set among the step S509 that discuss hereinafter.If in step S505, determine, on inspection DISPLAY ORDER all images ID in the formation is set, handle so and advance to step S510.
If in step S505, determine, the image I D that not every DISPLAY ORDER is provided with in the formation has passed through inspection, in other words, exists some DISPLAY ORDERs that the image I D that check mark is not set in the formation is set, handle so and advance to step S506, wherein CPU20 increases progressively 1 with frame counter.
Then in step S507, CPU 20 with reference to determine and be stored in omission cycle and frame counter value in the register, determine whether to be presented at the frame of representing in the frame counter.Or rather, be that 1/2 time frame counter is represented 2n (n is a positive integer) if work as the omission cycle, CPU20 determines to show topics frame so, in other cases, CPU 20 determines not show this theme frames.If working as the omission cycle is that 1/3 time frame counter is represented 3n (n is a positive integer), CPU 20 determines to show topics frames so, and in other cases, CPU 20 determines not show this theme frames.If determine that in step S507 the theme frame will show that promptly this frame does not omit, handle advancing among the step S509 so.
If in step S507, determine, do not show this frame, promptly omit this frame, handle advancing to step S508 so, wherein CPU 20 is provided with in the formation at DISPLAY ORDER and in the theme frame delete flag is set.
If in step S507, determine to show this frame, perhaps after step S508, handle advancing to step S509.In step S509, CPU 20 is provided with check mark in the image I D corresponding to the frame of representing in frame counter.Processing turns back to step S505, and repeats corresponding the processing.
If determine in step S501, playback speed is not fast, has determined on inspection in step S505 that perhaps DISPLAY ORDER is provided with all images ID in the formation, handles advancing to step S510 so.In step S510, CPU 20 is provided with the speed value of the setting GOP_Speed that is provided with in the formation when the execution omission handles 2 at DISPLAY ORDER.Processing turns back to the step S479 of Figure 40.
According to omit handling 2, before carrying out the decoding plan, be provided with at DISPLAY ORDER and want in the formation in the frame of abridged delete flag to be set, and can detect when planning when decoding will be at the image type of the image of codec processing unit stem demonstration.After this handles routine, the B image of only having decoded and will show, and will not offer decoding processor 77 by abridged B image.Want abridged I image or P image although decoded, also do not show them.
Be described in the frame control and treatment of carrying out among the step S435 among Figure 39 2 below with reference to Figure 42.
In step S541, CPU 20 is by reference time information I/P picture decoding formation and be used for managing the time management counter of the timing of each processing of carrying out at replay device 1, determines whether to exist over and done with any flow data of its demonstration time.If in step S541, determine to exist over and done with flow data of its demonstration time, handle advancing to step S547 so, wherein carry out the underflow of discussing with reference to Figure 35 and handle.Handle then and turn back to step S435 among Figure 39.
If determining in step S541 does not have over and done with flow data of its demonstration time, execution in step S542 is to step S545 so, and it is similar among Figure 18 step S192 respectively basically to step S195.
Or rather, in step S542, carry out I image/P picture decoding of discussing below with reference to Figure 19 and handle.In step S543, carry out the B picture decoding of discussing below with reference to Figure 20 and handle.In step S544, CPU 20 transmission displays image information are to CPU 11 and increase progressively the time management counter.To frame (not having the abridged frame) execution in step S542 that delete flag wherein is not set among the step S508 in Figure 41 to step S544.
Then, in step S546, carry out the omission of describing below with reference to Figure 43 and handle 3, and processing turns back to step S435 among Figure 39 then.
According to frame control and treatment 2,, send to CPU 11 so according to the decoding frame of planning to decode, and with displays image information, and carry out to omit and handle 3 if there is not its demonstration time flow data of process.If there is its demonstration time flow data of process, carries out the underflow of describing with reference to Figure 35 so and handle.
With reference now to the flow chart of Figure 43, is given in the description of the omission processing of carrying out among the step S546 among Figure 42 3.
In step S571, CPU 20 determines whether playback speed is handled (omit processing 2 if carry out in the plan processing 2 of decoding, then handle 2 for omission) time than unidirectional previous omission and wanted soon.Or rather, if present speed (speed) is greater than being provided with the GOP_speed that is provided with in the formation at DISPLAY ORDER, if promptly speed/GOP_speed>1 remains really, CPU 20 determines that playback speed becomes faster.If determine that in step S571 the speed that playback speed before omitted in the processing unlike same direction is fast, be that the playback speed reservation is identical or slower, the direction of perhaps resetting is oppositely, handles turning back among the step S546 of Figure 42 so, and advances among Figure 39 among the step S436.
If determine in step S571, the present speed speed of before having omitted in the processing than same direction that becomes is fast, handles so and advances to step S572.In step S572, CPU 20 determines the omission cycle according to the speed value of setting in present speed value of setting and the omission processing formerly 2, and stores the determined omission cycle.
Or rather, CPU 20 is provided with the omission cycle according to the inverse of the absolute value of speed/GOP speed.For example, if present speed change to from previous speed * 2 * 4, so CPU 20 just the omission cycle be set to 1/2, and it is stored in the register.
In step S573, CPU 20 resets and is used to count the frame counter of frame number, and no matter omitting the codec processing unit of handling in 3.
Whether in step S574, determine that check mark is provided with each image I D in the formation and is provided with relatively with being stored in the temporal information DISPLAY ORDER, thereby determined whether on inspection all images ID that is provided with in the formation to be set at the temporal information DISPLAY ORDER.Check mark is to be arranged on the temporal information DISPLAY ORDER among the step S579 that discusses hereinafter mark in the formation is set.If find that in step S574 inspected is stored in the temporal information DISPLAY ORDER all images ID in the formation is set, handle so and turn back to step S546 among Figure 42, and advance to step S436 among Figure 39.
If in step S574, determine, not to have checked at the temporal information DISPLAY ORDER all images ID in the formation is set, that is, exist some that the image I D of check mark also is not set, handle advancing to step S575 so, wherein CPU 20 increases progressively 1 with frame counter.
Then, in step S576, CPU 20 is with reference to determined and be stored in omission cycle and frame counter value in the register, to determine whether the being presented at frame of representing in the frame counter.Or rather, be that 1/2 time frame counter is represented 2n (n is a positive integer) if work as the omission cycle, CPU20 just determines that the theme frame will show so, in other cases, CPU 20 determines that the theme frames do not show.If working as the omission cycle is that 1/3 time frame counter is represented 3n (n is a positive integer), CPU20 just determines that the theme frame will show so, and in other cases, CPU 20 determines that the theme frames do not show.If determine that at step S576 the theme frame will show that promptly this frame does not omit, handle so and advance to step S579.
If determine that in step S576 the theme frame does not show that promptly this frame will omit, handle advancing to step S577 so.In step S577, CPU 20 is provided with the formation image I D of deletion theme frame from the temporal information DISPLAY ORDER, perhaps in the temporal information DISPLAY ORDER is provided with image I D in the formation mark is set and does not show (not show tags) to represent this frame.
Carrying out the temporal information replacement of describing with reference to Figure 24 in step S578 handles.In this case, the temporal information added value added_count that calculates in step S302 handles when changing from previous omission when speed to handle the picture number that changes by the omission among step S576 and S577.
If in step S576, determine, be presented at the frame of representing in the frame counter, perhaps after step S578, handle advancing to step S579.In step S579, CPU 20 is provided with check mark in the temporal information DISPLAY ORDER is provided with the image I D of the frame of representing in the formation in frame counter.Handle then and turn back to step S574 and repeat corresponding the processing.
Handle 3 according to omitting, if rapid change must be also faster than omitting the speed of handling in 2 in same direction, the image I D of the frame that deletion does not show the formation is set from the temporal information DISPLAY ORDER so, perhaps in the temporal information DISPLAY ORDER is provided with respective image ID in the formation, not show tags is set.B image demonstration processing is carried out in formation and a frame delay shows set handling by being provided with reference to the temporal information DISPLAY ORDER of resetting.If it is also faster than the speed of handling in 2 in unidirectional omission that speed becomes, change the temporal information DISPLAY ORDER according to the speed of setting so setting in the formation is set.Therefore, after this handled routine, in other words the B image of only having decoded and will show, abridged B image not offer decoding processor 77.Want abridged I image and P image although decoded, also it is not shown.
Therefore, even after finishing the decoding plan, for each frame determines whether to have changed the speed setting.If rapid change must be also faster than the speed of handling in 2 in unidirectional omission, the temporal information of resetting so DISPLAY ORDER is provided with formation, thereby carries out the decoding plan so that can change and the playback stream data by rapid response speed.
When a GOP comprises 15 width of cloth images, the above-mentioned decoding processing of having described.Described such situation below with reference to Figure 44 to 64, the number that wherein is included in an anchor frame (I image and P image) among the GOP is 6 or more or 4 or still less.
If to be similar to the mode that replay device 1 as shown in Figure 1 is the decoding processing that comprises that each GOP of 15 width of cloth images carries out, be that each GOP that comprises many anchor frames carries out decoding, shown in Figure 44 and 45, some images are not decoded or are postponed the decoding of some images.
With reference now to Figure 44, provides the description of this situation, be the mode of the decoding processing that comprises that a GOP of 15 width of cloth images carries out GOP that decodes to be similar to wherein by a decoder (in the decoder 22,23 and 24), the stream of the GOP (0) that comes forward playback to comprise to form with * 1 speed, the GOP (1) that forms by 21 width of cloth images, the GOP (2) that forms by 15 width of cloth images from B0 to P14 from B0 to P20 by 15 width of cloth images from B0 to P14.
Utilize previous GOP to decode at the B image of the stem of each GOP.Owing to store 7 width of cloth anchor frames among GOP (1), 6 groups of the anchor frame that is used in 8 groups in video group memory 82 to decode are just occupied by I2 to P17, and do not have the space to give the ensuing anchor frame P20 of GOP (1) and the I2 of GOP (2).Therefore, in the decoder of decoding GOP (1), during decoding, the I2 of the anchor frame P20 of GOP (1) and GOP (2) can not be used as reference picture, thereby can't decode the P17 of GOP (1) and B0 and the B1 of frame and GOP (2) subsequently.
With reference now to Figure 45, provides the description of following situation, be the mode of the decoding processing that comprises that a GOP of 15 width of cloth images carries out GOP that decodes by a decoder one of ( decoder 22,23 and 24) to be similar to wherein, coming to reset with *-1 velocity reversal comprises by from the GOP (0) of 15 width of cloth image constructions of B0 to P14, by from the GOP (1) of 21 width of cloth image constructions of B0 to P20 with by the stream from the GOP (2) of 15 width of cloth image constructions of B0 to P14.
Utilize GOP subsequently to decode at the B of each GOP stem image.Because as shown in figure 44,7 anchor frames are arranged among the GOP (1), in 8 groups in video group memory 82, occupy by I2 to P17 and to be used to 6 groups of anchor frame of decoding, and do not have the space to give the ensuing anchor frame P20 of GOP (1) and the I2 of previous GOP (2).Therefore, in the decoder of decoding GOP (1), during decoding, the I2 of the anchor frame P20 of GOP (1) and previous GOP (2) can not be used as reference picture, thereby can't decode the P17 of GOP (1) and B0 and the B1 of frame and GOP (2) subsequently.
As discussed above, owing to form a large amount of anchor frames of the image of a GOP, if be to comprise that the mode that each GOP of 15 width of cloth images carries out carries out decoding processing, some frames of not decoding so to be similar to by replay device 1 shown in Figure 1.
Below with reference to Figure 46 and 47 such situation is discussed, is wherein had 4 or still less the GOP of anchor frame and the GOP with 5 or more anchor frames constantly carry out quick replay operations one.
With reference now to Figure 46, provides the description of following situation, wherein by by a decoder one of ( decoder 22,23 and 24) to be similar at the mode that comprises the decoding processing of carrying out on the GOP of 15 width of cloth images GOP that decodes, the speed forward playback with * 2 comprises GOP (0), GOP (1) and GOP (2) (wherein each is to be formed by 3 width of cloth image B 0, B1 and I2) and comprises by the GOP (3) of 15 width of cloth image constructions from B0 to P14 and the stream of GOP (4).
Utilize the decode B image of each GOP stem of previous GOP.In GOP (0), GOP (1) and GOP (2), an anchor frame is only arranged, and among the GOP (3) 5 anchor frames are being arranged subsequently.Between the B0 and B1 of GOP (0), the B0 that in decoder 2, utilizes the previous frame decoding to show, then, the B1 that will show among the I2 of the GOP (0) that in decoder 0, decodes and the GOP (1), and the B0 that in decoder 1, decodes and to show among the I2 of GOP (1) and the GOP (2) then.In each decoder, do not begin all images in exporting decoding processing formerly of decoding processing subsequently.
That is to say,, thereby postpone to occur in the decoding processing because decoder 0 begins the anchor frame of GOP (3) of decoding after the B1 of output GOP (1).Therefore, keep output until the decoding that after the anchor frame of decoding GOP (3), finishes B3 (at first demonstration) by the B1 of the GOP (3) of decoder 2 decoding.
With reference now to Figure 47, provides the description of this situation, wherein by by a decoder one of ( decoder 22,23 and 24) to be similar at the mode that comprises the decoding processing of carrying out on the GOP of 15 width of cloth images GOP that decodes, come reverse reproduction to comprise that (wherein each is by 3 width of cloth image B 0 for GOP (4), GOP (3) and GOP (2) with *-2 speed, B1, I2 constitute) and comprise by from the GOP (1) of 15 width of cloth image constructions of B0 to P14 and the stream of GOP (0).
Utilize the decode B image of each GOP stem of previous GOP.In GOP (4), GOP (3) and GOP (2), an anchor frame is only arranged, and among the GOP (1) 5 anchor frames are being arranged subsequently.Between B0 and B1 prior to the GOP of GOP (4), the B0 that in decoder 1, utilizes the I2 of GOP (4) to decode and will show, then, between the B0 and B1 of GOP (4), in decoder 0 with the come together B1 that decodes and will show of the I2 of GOP (3).And then, between the B0 and B1 of GOP (3), the B1 that in decoder 2, decodes and will show with the I2 of GOP (2).Then, between the B0 and B1 of GOP (2), the B0 that in decoder 1, will show with GOP (1) decoding.In each decoder, do not begin decoding processing all images in the output early decoding is handled subsequently.
That is to say,, thereby postpone to occur in the decoding processing because decoder 1 begins the anchor frame of GOP (1) of decoding after the I2 of output GOP (4).Therefore, the I2 by the GOP (3) of decoder 2 decoding keeps output until the at first decoding of the B0 of demonstration that finishes GOP (2) after the anchor frame of decoding GOP (1).
Like this, if GOP that comprises minority anchor frame and the GOP execution that comprises most anchor frames are similar to the decoding processing that is comprised a GOP execution of 15 width of cloth images by 1 pair of replay device shown in Figure 1, needing so the expends time in anchor frame of decoding, thus delay in decoding processing caused.
In order to deal with such situation, Figure 44 to 47 is described as reference, and wherein the number of image is not 15 width of cloth in the replay device 1 shown in Figure 1 in GOP, can take following measure so.After the structure that has detected each GOP, can rebuild its decoding unit by dividing or merging GOP as codec processing unit, comprise anchor frame corresponding to the group number of the video group memory 82 of decoder 22,23 or 24.
Below with reference to Figure 48 to 50B division is discussed and is comprised the GOP of 30 frames and the GOP that decoding is divided.
GOP (0) is made of GOP (1) 30 frames before, as shown in figure 48, i.e. and B0 to P29.In each codec processing unit, with previous GOP preceding two the B images of decoding.
Among 8 groups of video group memory 82, can give the anchor frame with 6 set of dispense.Owing to 10 anchor frames are arranged, thereby not every frame can be by decoding with the identical decoding processing of carrying out on each GOP that comprises 15 width of cloth images in GOP (0).Therefore, shown in the bottom of Figure 48, GOP (0) is divided into two codec processing units, so that the number of anchor frame becomes 6 or still less, and then, carry out decoding so that also decoding be used for decoding all P images and the required anchor frame of B image of the GOP that is included in division.
GOP (0) is divided into two codec processing units, and shown in Figure 49 A and 49B, GOP (0-0) comprises I2 to P17, and GOP (0-1) comprises B18 to P29 and B0 and B1, and by different decoders decode GOP (0-0) and GOP (0-1).Need the last anchor frame P17 of GOP (0-0) to be used for two B image B 18 and the B19 of decoding in GOP (0-1) stem.That is to say that P17 also is used as the anchor frame of GOP (0-1).Need the anchor frame of GOP (0-0) be used to the to decode anchor frame of GOP (0-1).Therefore, after the anchor frame of the GOP (0-0) that decodes by the group of using 6 decoding anchor frames, anchor frame by using these 6 group (by the rewriting group) GOP that decode (0-1) and the I image of GOP subsequently.
Thereby, at the decoder of GOP (0-0) or GOP (0-1) that is used for decoding, decoding anchor frame in fixing group, and the direction of no matter resetting is forward or oppositely or no matter playback speed is normally or fast, do not change the decoding order of anchor frame.On the contrary, in GOP (0-0) and GOP (0-1), the decoding order of B image is according to playback direction or playback speed and different.In Figure 49 A and 49B, show * 1 and * decoding order of-1 forward and reverse playback speed.
In * 1 replay operations, shown in Figure 50 A, in a decoder, after the anchor frame I2 to P17 of decoding GOP (0-0), the B image B 3 of decoding GOP (0-0) is to B16.Simultaneously, in another decoder, after decoding anchor frame I2 to P14, by using anchor frame I2 to P14 as reference image decode the anchor frame P17 to P29 of GOP (0-1) and the I2 image of GOP (1), and the B image B 18 to B28 of decoding GOP (0-1) and the B0 to B1 of GOP (1) then.
In *-1 replay operations, shown in Figure 50 B, in a decoder, after decoding anchor frame I2 to P14, by using anchor frame I2 to P14 as reference image decode the anchor frame P17 to P29 of GOP (0-1) and the I2 frame of GOP (1), and then, the B image B 28 of the B1 of decoding GOP (1) and B0 and GOP (0-1) is to B18.In another decoder, after the anchor frame I2 to P17 of decoding GOP (0-0), the B image B 16 of decoding GOP (0-0) is to B3.
Next, with reference to Figure 51 to Figure 53 B setting and decoding processing when each GOP (each comprise 4 or still less anchor frame) consecutive hours codec processing unit are described.
Shown in Figure 51, GOP (0) is formed by 6 frame B0 to P5, and GOP (1) is formed by 9 frame B0 to P8, follows by GOP (2).In each codec processing unit, with previous GOP preceding two the B images of decoding.
Between 8 groups of video group memory 82, give the anchor frame with 6 set of dispense.Two anchor frames are arranged in GOP (0) and three anchor frames are arranged in GOP (1).Therefore,, merge GOP (0) and GOP (1), so that the number of anchor frame becomes 6 or still less as shown in Figure 51 bottom.
In this case, with previous GOP decode preceding two B frame B0 and the B1 of GOP (0), and as I2 to P5, the GOP (1) of same codec processing unit GOP (0-0) decoding GOP (0) and preceding two B image B 0 and the B1 of GOP (2).Decoding anchor frame in fixing group, and the decoding order of anchor frame do not change, shown in Figure 52, the direction of no matter resetting is a forward or oppositely or no matter playback speed is normally or fast.On the contrary, the decoding order of B image is according to playback direction or playback speed and difference.In Figure 52, be illustrated in * 1 or * decoding order of-1 forward or reverse playback speed.
In the operation of * 1 forward playback, shown in Figure 53 A, in any decoder, the anchor frame of decoding GOP (0-0) and the I2 frame of GOP (2), and the B image of the GOP (0-0) that decodes then.Then, in *-1 reverse reproduction operation, shown in Figure 53 B, in decoder whichsoever, be to carry out the reverse reproduction operation, the anchor frame of decoding GOP (0-0) and the I2 frame of GOP (2), and then, the B image of decoding GOP (0-0).
May there be such situation, wherein comprises 6 or the GOP of more anchor frames and have 4 or the GOP of anchor frame still less continuously.In this case, describe by merging this GOP below with reference to Figure 54 to 56B and repartition them then and come equably distribution load to the setting of the codec processing unit of decoder.
Shown in Figure 54, GOP (0) comprises 21 frame B0 to P20, and GOP (1) comprises 9 frame B0 to P8, follows by GOP (2).In each codec processing unit, with previous GOP preceding two the B images of decoding.
Among 8 groups of video group memory 82, can give the anchor frame with 6 set of dispense.Yet, because 7 anchor frames are arranged in GOP (0), if thereby GOP (0) carried out decoding to be similar to when a GOP comprises the mode of 15 width of cloth images, so wrong the generation.In this case, GOP (0) can be divided into two codec processing units so that the number of anchor frame becomes 6 or still less.Yet, because the number of the frame of GOP is little subsequently, thereby shown in Figure 54 bottom, merge and repartition GOP (0) and GOP (1), so that, carry out decoding according to these codec processing units then from GOP (0) and GOP (1) generation two codec processing unit GOP (0-0) and GOP (0-1).
Merging GOP (0) and GOP (1) afterwards, shown in Figure 55 A, the I2 to P17 of GOP (0) is set to the first codec processing unit GOP (0-0), and shown in Figure 55 B, the B1 of the B18 to P20 of GOP (0), whole GOP (1) and GOP (2) and B2 are set to the second codec processing unit GOP (0-1), and first and second codec processing units of decoding in different decoders.The last anchor frame P17 of GOP (0-0) is that be used to decode preceding two B image B 18 and the B19 of GOP (0-1) is required.That is to say that P17 also is used as the anchor frame of GOP (0-1).Need the anchor frame of GOP (0-0) be used to the to decode anchor frame of GOP (0-1).Therefore, after the anchor frame of the 6 groups of GOP that decode (0-0) that use decoding anchor frame, anchor frame by using these 6 groups (by rewriteeing this 6 groups) GOP that decode (0-1) and the I image of GOP subsequently.
Therefore, at the decoder of GOP (0-0) or decoding GOP (0-1) that is used for decoding, decoding anchor frame in fixing group, and no matter the playback direction is a forward or oppositely or no matter playback speed is normally or fast, the decoding order of anchor frame does not change.On the contrary, in GOP (0-0) and GOP (0-1), the decoding order of B image is according to playback direction or playback speed and difference.In Figure 55 A and 55B, illustrated * 1 or * decoding order of-1 forward or reverse playback speed.
Or rather, in * 1 forward playback operation, shown in Figure 56 A, in a decoder, at the I2 to P17 (it is the anchor frame of GOP (0-0)) of decoding GOP (0) afterwards, the B3 to B16 (it is the B image of GOP (0-0)) of decoding GOP (0).In another decoder, after the anchor frame I2 to P14 of decoding GOP (0-0), by using I2 to P14 as reference image decode P17 and I2, the P5 of P20 and GOP (1) and the I2 of P8 (it is the anchor frame of GOP (0-1)) and GOP (2) of GOP (0), then, B18 and the B0 to B7 (it is the B image of GOP (0-1)) of B19 and GOP (1) and B0 and the B1 of GOP (2) of decoding GOP (0).
In *-1 reverse reproduction operation, shown in Figure 56 B, in a decoder, after the anchor frame I2 to P14 of decoding GOP (0-0), by using I2 to P14 as reference image decode P14 and I2, the P5 of P20 and GOP (1) and the I2 of P8 (it is the anchor frame of GOP (0-1)) and GOP (2) of GOP (0).Then, B19 and the B18 (it is the B image of GOP (0-1)) of the B7 to B0 of the B1 of decoding GOP (2) and B0 and GOP (1) and GOP (0).In another decoder, at decoding I2 to P17 (it is the anchor frame of GOP (0-0)) afterwards, the B16 to B3 (it is the B image of GOP (0-0)) of decoding GOP (0).
May have such situation, wherein comprise continuously 4 or still less the anchor frame GOP and have the GOP of 6 or more anchor frames.In this case, describe by merging that these GOP repartition them then so that equably sharing of load is given the setting of the codec processing unit of decoder below with reference to Figure 57 to 59B.
Shown in Figure 57, GOP (0) comprises 9 frame B0 to P8, and GOP (1) comprises 21 frame B0 to P20, follows by GOP (2).In each codec processing unit, with previous GOP preceding two the B images of decoding.
Among 8 groups of video group memory 82, can give the anchor frame with 6 set of dispense.Although in GOP (0), have only three anchor frames, 7 anchor frames are arranged in GOP (1), therefore and merge and repartitioned GOP (0) and GOP (1), shown in the bottom of Figure 57, so that generate two codec processing unit GOP (0-0) and GOP (0-1) from GOP (0) and GOP (1), carry out decoding according to these codec processing units then.
Merging GOP (0) and GOP (1) afterwards, shown in Figure 58 A, the P8 of the I2 to GOP (1) of GOP (0) is set to the first codec processing unit GOP (0-0), and shown in Figure 58 B, the B1 of the B9 to P20 of GOP (1) and GOP (2) and B2 are set to the second codec processing unit GOP (0-1), and first and second codec processing units of decoding in different decoders.The last anchor frame P8 of GOP (0-0) is that be used to decode preceding two B image B 9 and the B10 of GOP (0-1) is required.That is to say that P8 also is used as the anchor frame of GOP (0-1).In the anchor frame of GOP (0-0), need the anchor frame I2 of GOP (1) and P5 be used to the to decode anchor frame of GOP (0-1).Therefore, after the I2 and P5 of the one group of GOP that decodes (1) in using 6 groups of decoding anchor frame, anchor frame by using these 6 groups (by rewriteeing this 6 groups) GOP that decode (0-1) and the I image of GOP subsequently.
Therefore, at the decoder of GOP (0-0) or decoding GOP (0-1) that is used for decoding, decoding anchor frame in fixing group, and no matter the playback direction is a forward or oppositely or no matter playback speed is normally or fast, the decoding order of anchor frame does not change.On the contrary, in GOP (0-0) and GOP (0-1), the decoding order of B image is according to playback direction or playback speed and difference.In Figure 58 A and 58B, illustrated * 1 and * decoding order of-1 forward and reverse playback speed.
Or rather, in the operation of * 1 forward playback, shown in Figure 59 A, in a decoder, at the I2 to P8 of decoding GOP (0) and the I2 to P8 (it is the anchor frame of GOP (0-0)) of GOP (1) afterwards, the B3 to B7 of decoding GOP (0) and the B0 to B7 (it is the B image of GOP (0-0)) of GOP (1).In another decoder, the decoding GOP (1) anchor frame I2 to P5 (it is the anchor frame of GOP (0-0)) afterwards, by using I2 to P5 as reference image the decode P8 to P20 (it is the anchor frame of GOP (0-1)) of GOP (1) and the I2 of GOP (2), then, the B9 to B19 (it is the B image of GOP (0-1)) of decoding GOP (1) and B0 and the B1 of GOP (2).
In *-1 reverse reproduction operation, shown in Figure 59 B, in a decoder, at the anchor frame I2 of decoding GOP (1) and P5 (it is the anchor frame of GOP (0-0)) afterwards, by using I2 and P5 as reference image the decode P8 to P20 (it is the anchor frame of GOP (0-1)) of GOP (1) and the I2 of GOP (2).Then, the B1 of decoding GOP (2) and the B19 to B9 (it is the B image of GOP (0-1)) of B0 and GOP (1).In another decoder, at the I2 to P8 of decoding GOP (0) and the I2 to P8 (it is the anchor frame of GOP (0-0)) of GOP (1) afterwards, the B7 to B0 of decoding GOP (1) and the B7 to B3 (it is the B image of GOP (0-0)) of GOP (0).
Figure 48 to 59B is described as reference, makes up decoding unit as codec processing unit.Then, be the plan of decoding of each decoding unit.And carry out the frame omission as required so that can carry out replay operations by the playback speed of specifying playback direction and appointment.Then, carry out the decoding and the also output stream of resetting.
Or rather, between a plurality of GOP that transmit and be stored in the memory 18, CPU 20 reads the GOP that will decode and the structure of GOP subsequently, makes up decoding unit by dividing based on the number of anchor frame and merging GOP then, and it is used as codec processing unit.CPU 20 transmitting control commands to PCI bridge 17 reads the flow data of each decoding unit with control PCI bridge 17 from memory 18 then, and the flow data that reads is offered in decoder 22,23 and 24 one.
Then, CPU 20 executive plan as mentioned above.In this case, the situation of being discussed as reference Figure 48,54 or 57 in GOP (0-1), anchor frame for decoding corresponding decoding unit, decode if desired another decoding unit to small part anchor frame, then the anchor frame of this decoding unit also is set in the input picture formation, from the stem of I/P picture decoding formation these frames is set in proper order.Yet, owing to do not show the required anchor frame of anchor frame of another decoding unit that is not used to decode, thereby it is not arranged on DISPLAY ORDER and is provided with in the formation.
CPU 20 decodes to carry out by the register value control decoder 22,23 or 24 of the decoder of referential expression reception subsequent data then. Decoder 22,23 or 24 basic stream address determining unit 73 are allowing Memory Controller 74 to read view data corresponding to the image I D that is provided with from input buffer 75 the formation of temporal information I/P picture decoding under the control of controller 20, and reading of data is supplied with decoding processor 77.
If the image that will decode is the I image, the I image supplied with from Memory Controller 74 of CPU20 control decoding processor 77 decodings so, also control write image address determining unit 78 with the frame data of decoding supply with Memory Controller 81 and with its be stored in video group memory 82 in the specified group of storage I image.If the image of decoding is the P image, CPU 20 control reference picture address determining units 79 are to allow Memory Controller 81 to read the reference picture that is stored in the video group memory 82 according to the reference group position of P image so, and the reference picture that reads offered decoding processor 77, and the P image supplied with from Memory Controller 74 of control decoding processor 77 decodings.CPU 20 also controls and writes image address determining unit 78 serves as in the group of storage P image and appointment with what the decoded frame image is supplied with Memory Controller 81 and it is stored in video group memory 82.If the image of decoding is the B image, CPU 20 control reference picture address determining units 79 are to allow Memory Controller 81 to read the reference picture that is stored in the video group memory 82 according to the reference group position of B image so, and will read reference picture and offer decoding processor 77, and the B image that provides from Memory Controller 74 with decoding of control decoding processor 77.
The GOP division/merging of flow chart description when the picture number that forms GOP is different from 15 below with reference to Figure 60 handled.
In step S601, CPU 20 determines whether also not have after having division GOP the decoding unit of plan.
If in step S601, find planless decoding unit, handle so and advancing to step S602, wherein CPU20 by carry out frame shown in Figure 4 handle 1 or frame shown in Figure 39 handle 2 and carry out the plan of planless decoding unit, and this decoding unit of decoding.
If in step S601, determine not have planless decoding unit, handle so and advance to step S603, wherein CPU 20 detects GOP and the structure of next planless GOP subsequently again.
In step S604, CPU 20 determines that subsequently GOP are whether greater than can be by the full-size of a decoder decode.Or rather, CPU 20 checks the number of the anchor frame of GOP subsequently and the group number of the anchor frame that is used in the inspection number of anchor frame and the decoder to decode is compared, to determine that whether GOP is greater than full-size subsequently.
If determine in step S604, GOP is greater than can be by the full-size of a decoder decode subsequently, handle so advance to step S605 with determine subsequently GOP and the merging size of more next GOP whether be can be by the size of two decoder decodes.Or rather, if video group memory 82 has 8 groups, whether CPU 20 to determine to merge the number of the anchor frame of GOP be 10 or still less so,, deducts the twice of 3 numbers that obtain or still less by the group number from video group memory 82 that is.
If the size that determine to merge in step S605 is can be by the suitable size of two decoder decodes (for example, when as described with reference to Figure 54 to 56B, have the long GOP of 6 or more anchor frames continuously and have 4 or still less during the short GOP of anchor frame), processing advancing to step S610.
If determine that at step S605 the GOP that is merged can be handled and be advanced to step S606 so by the size of two decoder decodes.At step S606, CPU 20 subsequently GOP be divided into a plurality of can be by the decoding unit of a decoder decode, in the long GOP that is discussed at reference Figure 48 to 50B with 30 width of cloth images.Handle then and advance to step S612.
If in step S604, determine, be not more than can be by the full-size of a decoder decode for GOP subsequently, handle advancing to step S607 so, with determine subsequently GOP and the merging size of more next GOP whether be can be by the suitable size of a decoder decode.
If determine in step S607, the size of merging is to be handled so and to be advanced to step S608 by the size of a decoder decode.In step S608, will be subsequently GOP and next again GOP merge into a decoding unit, have 4 or the situation of the GOP of anchor frame still less as continuous each, describe as reference Figure 51 to 53B.Handle then and advance to step S612.
If determine in step S607, the size of merging can not become and can handle advancing to step S609 to determine whether the size that is merged can be by the size of two decoder decodes by the size of a decoder decode.Or rather, Figure 57 to 59B discusses as reference, if have continuously 4 or still less the anchor frame short GOP and have the long GOP of 6 or more anchor frames, if and video group memory has 8 groups, CPU 20 anchor frame number 10 frames or still less whether of determining the GOP that merge so, that is, deduct the twice of 3 numbers that obtain or still less by the group number from video group memory 82.
If definite size that merges becomes and can handle advancing to step S610 by the size of two decoder decodes in step S605 or S609.In step S610, CPU 20 merges two GOP and then it is reclassified as two decoding units, discusses as reference Figure 54 to 56B or Figure 57 to 59B.Handle then and advance to step S612.
If the size that determine to merge at step S609 can not become and can handle advancing to step S611 by the size of two decoder decodes, wherein CPU 20 subsequently GOP be set to a decoding unit.
After step S606, S608, S610 or S611, handle to advance to step S612, wherein CPU20 by carry out frame shown in Figure 4 handle 1 or frame shown in Figure 39 handle 2 and plan the decoding of decoding unit subsequently, and this decoding unit of decoding.
Then, in step S613, if there is the decoding unit of planless division, CPU 20 is stored in it and waits in the tabulation so.Finish processing then.
Handle according to above-mentioned GOP division/merging, even even form the picture number of GOP be not 15 or picture number according to GOP and difference, also can carry out the division of GOP or merge to reach the fast decoding processing.
In above-mentioned example, when having described video group memory 82 when decoder 22,23 or 24 and having 8 groups and when 6 groups were fixed for decoding the anchor frame in 8 groups, the division of GOP or merging were so that reach effective plan.Even be stored in the number of the group (frame) in the video group memory 82 and be not 8 or the fixedly group number that promptly is used in decoding anchor frame be not 6, also can be according to the group number in the video group memory 82 or the fixedly group number of the anchor frame that is used to decode suitably carry out division or the merging of GOP.More precisely, if the fixedly group number of the anchor frame that is used to decode is provided with decoding unit so and becomes less than 6 so that be included in the number of an anchor frame in the decoding unit less than 6.On the contrary, being included in the number of an anchor frame in the decoding unit if the fixedly group number of the anchor frame that is used to decode greater than 6, is provided with decoding unit so becomes less than the group number.
Handle according to GOP division/merging, shown in Figure 61, can in * 1 replay operations, carry out decoding processing no problemly the GOP of 21 frames that comprises shown in Figure 44.
Or rather, the GOP (1) that will comprise 21 frames is divided into then two decoding units by decoder 1 and 2 decodings, that is, and and first decoding unit and second decoding unit.In this case, in decoding not from the anchor frame of first decoding unit of decoder 2 output so that after the anchor frame of second decoding unit of decoding, by using anchor frame as the fixing group of the anchor frame anchor frame of rewriting first decoding unit (as required by) second decoding unit of decoding.
Similarly, shown in Figure 62, can in *-1 replay operations, carry out decoding processing without any problem ground to the GOP of 21 frames that comprises shown in Figure 45.
Or rather, will comprise that the GOP (1) of 21 frames is divided into two decoding units, that is, first decoding unit and second decoding unit, it is subsequently by decoder 2 and decoder 1 decoding.In this case, in decoding not from the anchor frame of first decoding unit of decoder 2 output so that after the anchor frame of second decoding unit of decoding, by using anchor frame as the fixing group of the anchor frame anchor frame of rewriting first decoding unit (as required by) second decoding unit of decoding.
Similarly, shown in Figure 63, in * 2 replay operations,, can carry out decoding processing as the GOP that has minority anchor frame as shown in figure 46 continuously with when having the GOP of most anchor frames no problemly.
Or rather, GOP (0) and the GOP (1) that respectively comprises B0, B1 and I2 merged in the decoding unit, and will comprise the GOP (2) of B0, B1 and I2 and comprise from GOP (3) merging of 15 frames of B0 to P14 and repartitioning then.
Between the B0 and B1 (B0 and the B1 of GOP (0)) of the decoding unit of the merging of GOP (0) and GOP (1), in decoder 2 with previous GOP or early decoding unit the decode B (0) that will show and the required reference picture I2 of B (0) that is used to decode.Between the B0 and B1 (B0 and the B1 of GOP (2)) of the merging decoding unit of GOP (2) and GOP (3) the first half, in decoder 0 with the merging decoding unit of GOP (0) and GOP (1) the decode B (0) that will show and the required reference picture I2 of B (0) that is used to decode.With the first half the merging decoding unit of GOP (2) and GOP (3) B (7) that decodes, it is the stem of the B image that will show of the second half the decoding unit of GOP (3) in decoder 1.Between the B0 and B1 of GOP (4), the B0 and the I2 (it is as the required reference picture of decoding B0) that in decoder 2, decode and will show with the second half the decoding unit of GOP (3).The subsequent picture of decoding I2 and GOP (4) in decoder 0.
Similarly, in *-2 replay operations,, shown in Figure 64, can carry out decoding processing without any problem ground as the GOP that has minority anchor frame as shown in figure 47 continuously with when having the GOP of most anchor frames.
Or rather, GOP (4) and GOP (3) (respectively comprising B0, B1 and I2) are merged in the decoding unit, and will comprise the GOP (2) of B0, B1 and I2 and comprise merging and repartition then from the GOP (1) of 15 frames of B0 to P14.
The image except B0 and B1 (B0 and the B1 of GOP (3)), formerly the B image that will show (B0 in Figure 64) between two B images and the I2 in the decoding unit (its be used to decode the required reference picture of B (0)) of merging decoding unit of GOP (4) and GOP (3) decodes in decoder 1.Between the B0 and B1 (B0 and the B1 of GOP (3)) of the merging decoding unit of GOP (4) and GOP (3), the B (0) and the I2 (it is the required reference picture of decoding B (0)) that in decoder 0, decode and will show with the second half the merging decoding unit of GOP (2) and GOP (1).Between the B9 and B10 (B9 and the B10 of GOP (1)) of the second half the merging decoding unit of GOP (2) and GOP (1), in decoder 2, be B9 and P8 and the P11 (it is the required reference picture of decoding B9) that the first half decodings of GOP (1) will show with decoding unit subsequently.With GOP (0) B (1) that decodes, it is the stem of the B image that will show of the first half the decoding unit of GOP (1) in decoder 1.
As mentioned above, suitably to divide according to the number of anchor frame or merge GOP be the decoding unit that the group number form of the video group memory 82 that provides of decoder 22,23 or 24 becomes to have some anchor frames with the basis to CPU 20.CPU 20 control PCI bridges 17 offer decoder 22,23 or 24 with the stream with each decoding unit then, and carry out the decoding plan or the Showplan of each decoding unit, and control decoder 22,23 or 24 then to carry out decoding.Utilize this configuration, though the number of anchor frame according to GOP and difference, replay device 1 also can be carried out fast decoding by all frames of decoding reliably and handle, and can not cause any delay in the decoding processing.
After dividing GOP, in second decoding unit, promptly, in decoding unit with the anchor frame that should after the anchor frame of decoding early decoding unit, decode, if the anchor frame number of second decoding unit is 6, if the stem of decoding unit is I image or P image so, beginning demonstration regularly with respect to the decoding of anchor frame of decoding second decoding unit, to begin skew regularly be 6 width of cloth images, if and stem is the B image, this skew is exactly 7 width of cloth images.If the number of anchor frame is n frame (n be 5 or still less) in second decoding unit, beginning skew regularly with respect to decoding beginning demonstration regularly so changes according to the anchor frame number, that is, be exactly the n frame so if stem is I image or P image, be exactly the n+1 frame so if stem is the B image.
In the foregoing description, to be set to fixedly group number with the anchor frame that is used to decode identical or still less for the number of the anchor frame that comprises in decoding unit.More preferably, the number that the anchor frame is set by the number of the anchor frame of theme decoding unit is compared with the anchor frame number of early decoding unit so as it can be than the anchor frame number of early decoding unit big m frame (m is the integer greater than 1, this can be provided with by test or experience, and more preferably is 2 to 4).
For example, if having the GOP and the GOP with 10 width of cloth anchor frames of an anchor frame continuously, the anchor frame sum of two GOP is exactly 11 so, and is difficult to form two decoding units from these two GOP.Therefore, decode so that " 1 anchor frame+1 ", " 5 anchor frame+1 " and " 5 anchor frame+1 ", thereby cause delay in decoding processing.
Thereby the number that the anchor frame is set by the anchor frame number of decoding unit is compared with the anchor frame number of early decoding unit is so that it is unlike the anchor frame number big 2 of early decoding unit or more.In this case, when the anchor frame sum of two GOP is 11, decode so that " 1 anchor frame+1 ", " 3 anchor frame+1 ", " 3 anchor frame+1 " and " 4 anchor frame+1 ".Thereby, in decoding processing, do not postpone.
In addition, (no matter having carried out above-mentioned two omits the processing routine or carried out only omission routine) can provide following label sets as the used metadata of the video data that is stored in the compressed encoding among the HDD16 in above-mentioned processing.Read label sets and represent to be stored in the video data of the compressed encoding among the HDD16 as whether effective from the data that HDD16 reads, whether decoding label sets presentation code data are effective as the data that will decode according to the decoding plan, and whether the encoded data conduct of show tags set representations wants data presented effective according to Showplan.Then, according to playback speed or the automatic update mark collection of playback direction, thereby improve planning management.
In this case, a series of plan and label sets that are used for previous variable velocity replay operations can be managed as metadata (register information) respectively.As required, can in the video data of compressed encoding, embed this metadata, or in recording medium, write down it such as HDD16 as grammer.
The number of decoder, the number of group, decoder ID etc. can be used as metadata (structure register information) and manage.In addition, playback speed and playback direction can be used as metadata (playback register information) and manage.In this case, metadata can be used as grammer and is embedded in the video data of compressed encoding, perhaps can record in the recording medium such as HDD16.
By reference metadata (register information), the plan in the past of can reusing is handled, thereby feasible plan is handled more rapidly and be more accurate.
Can be as the above-mentioned metadata of data base administration in external equipment.
In the above-described embodiments, decoder 22,23 or 24 unnecessary complete decoding are stored in the video data of the compressed coding among the HDD16.
Or rather, in embodiments of the present invention, decoder 22,23 or 24 can only be decoded and de-quantization variable-length codes and need not to carry out inverse discrete cosine transformation (IDCT), and perhaps decoder 22,23 or 24 can the de-quantization variable-length codes and be need not its decoding.In this case, decoder 22,23 or 24 can generate register information, and its presentation code or decoding processing have been carried out which degree (for example, to de-quantization), and output and the incomplete relevant register information of decoded data.
In addition, in the aforementioned embodiment, incomplete coded data (for example, the data of not passing through variable length code through DCT and quantification) and the register information (if desired) that relates to the Code And Decode processing can be stored among the HDD 16.In this case, can under the control of CPU 20, decode incomplete coded data and be converted into baseband signal of decoder 22,23 or 24.
Or rather, decoder 22,23 or 24 can not carry out IDCT and de-quantization to incomplete coded data (it passes through variable length code through dct transform and quantification), and need not the code of decoding variable length.
In this case, the register information that CPU 20 can obtain to be stored in is among the HDD 16, relevant with incomplete coded data, handle about Code And Decode, and according to the register information that the is obtained plan of decoding.
In the aforementioned embodiment, not exclusively coded data and the register information (as required) handled about Code And Decode can be stored among the HDD 16, and decoder 22,23 or 24 unnecessary under the control of CPU 20 this incomplete coded data of complete decoding.
Equally in this case, the register information that for example, that CPU 20 can obtain to be stored in is among the HDD 16, relevant with incomplete coded data, handle about Code And Decode, and according to the register information that the is obtained plan of decoding.Decoder 22,23 or 24 also can generate the register information of handling about Code And Decode, and itself and incomplete decoded data are exported relatively.
In other words, decoder 22,23 or 24 can decode on ground, the control lower part of CPU 20 (operating part decoding processing step).That CPU 20 can obtain to be stored in is among the HDD 16, that be relevant to incomplete coded data, about the register information of the processing of Code And Decode, and carry out plan by decoder 22,23 or 24 decodings of carrying out according to the register information that is obtained.As required, decoder 22,23 or 24 can generate the register information of handling about Code And Decode, and output and the incomplete relevant register information of decoded data.
In HDD 16, also can store relatively about the register information that Code And Decode is handled, and CPU 20 can carry out the decoding plan according to register information with the compressed encoding flow data.Even when decoder 22,23 or 24 can be under the control of CPU 20 decoding compressed encoding stream data and be converted into baseband signal, also can generate the register information of handling about Code And Decode as required, and this information is exported on dependency basis band signal ground.
Although in the above-described embodiments, replay device 1 comprises a plurality of decoders, also decoder and replay device 1 can be provided discretely.
In this case, each independently decoder can receive the decode the video data of compressed coding, and show and the output result data.In addition, as mentioned above, each decoder can receive the video data with the compressed coding of partial decoding of h, and exports data through decoding to external source with the register information of handling about Code And Decode.Alternatively, each decoder can receive the decode the part coded data and be converted into baseband signal, and it is outputed to external source.Alternatively, each decoder can receive with the video data of partial decoding of h part compressed encoding and with decoded data and output to external source with the register information of handling about Code And Decode.
In the above-described embodiments, CPU 11 and CPU 20 separately provide.Yet CPU 11 and CPU20 can be integrated among the CPU, so that control whole replay device 1, even perhaps CPU 11 separates structure with CPU 20, but also it can be integrated on the chip.
Provide if CPU 11 separates with CPU20, the section processes of being carried out by CPU 11 in the previous embodiment at least can be carried out with time-sharing format by CPU 20 so.Alternatively, the section processes of being carried out by CPU 20 can be carried out with time-sharing format by CPU 11.That is to say that the processor that can carry out distributed treatment can be used as CPU 11 and CPU 20.
Alternatively, replay device 1 can constitute with network and link to each other, and can be carried out by the CPU of another device that is connected to network by the processing that CPU 11 or CPU 20 carries out to small part in the above-described embodiments.
Similarly, although in the above-described embodiments, memory 13 separates with memory 21 to be provided, and also can be integrated in the memory in the replay device 1.
In the above-described embodiments, HDD 16, decoder 22,23 and 24 and selector 25 be connected with each other so that it is integrated in the replay device 1 via bridge and bus.Yet the some of them element can externally connect by wired or wireless mode, and perhaps these devices can be connected to each other with another kind of connected mode.
Although in the above-described embodiments, compressed flow data is stored among the HDD 16, and it also can be stored in another recording medium, such as CD, magneto optical disk, semiconductor memory or disk.
CPU 20, memory 21, memory 18, decoder 22,23 and 24 and selector 25 be installed on the same expansion card, for example, pci card or PCI-express card.Yet if for example the transfer rate between the PCI-express technology card is very high by using, these elements also can be installed on the different expansion cards discretely so.
Can use the coding method of for example H264/AVC that is different from MPEG.
Can carry out a series of above-mentioned works of treatment by hardware or software.If use software, so with the corresponding software program from recording medium be installed to the computer that is structured in the specialized hardware or such as personal computer pass through wherein install in the computer that various programs can carry out various functions.In this case, replay device 1 shown in Figure 1 can be formed by for example personal computer shown in Figure 65 201.
In personal computer shown in Figure 65 201, CPU (CPU) 211 is according to being stored in the program in the read-only memory (ROM) 212 or carrying out various works of treatment from the program that memory cell 218 is loaded into the random access storage device (RAM) 213.In RAM213, also store CPU 211 and carry out the processing desired data.
CPU 211, ROM 212 and RAM213 interconnect each other via bus 214.Input/output interface 215 also can be connected on the bus 214.
Input unit 216, such as keyboard and mouse, output unit 217, such as display or loud speaker, memory cell 218 all is connected to input/output interface 215 such as hard disk and communication unit 219 such as modulator-demodulator or terminal adapter.Communication unit 219 such as the internet, is held communication via network.
Driver 220 also is connected to input/output interface 215 as required, disk 231, CD 232, magneto optical disk 233 or semiconductor memory 234 are installed in the driver 220 as required, and are installed to the memory cell 218 from the computer program that this recording medium reads.
As mentioned above, if software is used to carry out a series of works of treatment, corresponding software programs can be installed to the computer via network or from recording medium so.
This recording medium can be wherein stored program packaged media, be located away from the distributed program that provides in device ground to the user, such as disk 231 (comprising floppy disk), CD 232 (comprising compact disc-ROM (CD-ROM) or digital multipotency CD (DVD)), magneto optical disk 233 (comprising mini disk (MD) (registered trade mark)), perhaps semiconductor memory 234.Alternatively, recording medium can be ROM212 or be included in hard disk in the memory cell 218 of wherein having stored program, provide it to the user in the time of wherein in being structured in device.
In this manual, the step that forms the program in the recording medium that is recorded in comprises the processing of carrying out with chronological order as described in the specification, and also can comprise processing parallel or that carry out separately.
One skilled in the art will appreciate that as long as in the scope of claims and equivalent thereof, can carry out various modifications, merging, son merging and conversion according to design needs and other factors.

Claims (28)

1. messaging device that is used for the video data of decoding compression-coded comprises:
At least one decoding device is used for the video data of decoding compression-coded;
Apparatus for controlling of supply, the video data that is used to control compressed coding is supplied with decoding device; And
Control device is used to control the processing of being carried out by apparatus for controlling of supply and decoding device,
Wherein control device is determined the order of the decoding processing carried out by decoding device, so that among the image that the codec processing unit of the decoding processing that is used for being carried out by decoding device comprises, and decoding I image and P image before the B image, and
According to the playback speed instruction that the playback speed command device provides, selecting among the image of control device from be included in codec processing unit will be from the image of decoding device output;
Wherein said decoding device comprises:
The decoding processing final controlling element is used to carry out decoding processing, and
The image apparatus for controlling of supply is used for control image is supplied with the decoding processing final controlling element,
The image apparatus for controlling of supply under the control of control device, among the image in being included in codec processing unit, with I image and P image and select to be supplied to the decoding processing final controlling element as the B image of the image of exporting from decoding device, and
The decoding processing final controlling element under the control of control device, the image that decoding is supplied with from the image apparatus for controlling of supply, and also output as will be from the image of decoding device output and selecteed B image.
2. messaging device according to claim 1 wherein provides a plurality of decoding devices, and this messaging device also comprises:
Output switch apparatus is used to receive the not packed data from a plurality of decoding device outputs, and is used for optionally exporting the not packed data that is received,
Wherein control device is also controlled the processing of being carried out by output switch apparatus.
3. messaging device according to claim 1, wherein decoding device comprises and is used to store the group memory of decoded picture, this group memory has the group that is less than the number that is included in the image in the codec processing unit, and
Decoding device is carried out decoding processing by the decoded image that use is stored in the group memory as the reference image as required.
4. messaging device according to claim 3, wherein organize picture number that memory can store than be included in I image in the codec processing unit and the number of P image big by 2.
5. messaging device according to claim 3, wherein the number of decoding device is 3, and each decoding device comprises the group memory that can store 8 width of cloth images.
6. messaging device according to claim 3, wherein control device control decoding device is to be stored in the fixed position that is used for I image and P image in the group memory with I image and P image.
7. messaging device according to claim 1, wherein, when the image of the stem of codec processing unit is I image among the decoded picture of decoding device output or P image, control device control decoding device in case the demonstration output timing that the decoding that decoding device begins to decode begins regularly and decoding device begins the output decoder image each other phase deviation first pre-determine the image of number.
8. according to the messaging device of claim 7, wherein first pre-determine number total big by 1 than I image and P image.
9. messaging device according to claim 1, wherein, when the image in the codec processing unit stem is B image among the decoded picture of decoding device output, control device control decoding device in case the demonstration output timing that the decoding that decoding device begins to decode begins regularly and decoding device begins the output decoder image each other phase deviation second pre-determine the image of number.
10. messaging device according to claim 9, wherein, the second predetermined number is total bigger by 2 than I image and P image.
11. messaging device according to claim 1, wherein when the playback speed of playback speed command device indication and playback direction are normal and forward respectively, control device control decoding device in case the demonstration that the decoding that decoding device begins to decode begins regularly and decoding device begins the output decoder image export timing each other phase deviation first pre-determine the image of number.
12. messaging device according to claim 11, the wherein first predetermined number total big by 1 than I image and P image.
13. messaging device according to claim 1, wherein when the heavily effect speed of playback speed command device indication and playback direction when being respectively normal and reverse, control device control decoding device in case the demonstration that the decoding that decoding device begins to decode begins regularly and decoding device begins the output decoder image export timing each other phase deviation second pre-determine the image of number.
14. messaging device according to claim 13, the wherein second predetermined number total big by 2 than I image and P image.
15. messaging device according to claim 1, wherein control device is determined the timing of decoding device execution decoding processing, if so that exported the timing piece image of decoded B image by the timing ratio of decoding device decoding B image.
16. messaging device according to claim 3, wherein the video data of compressed encoding comprises image sets, and
When reception comprises the image sets of I image or P image, the number of wherein said image deducts 3 values that obtain greater than the group number from the group memory, control device is divided into codec processing unit with image sets, each codec processing unit comprises number than deducting I image or the P image that 2 values that obtain are lacked by the group number from the group memory, and the control apparatus for controlling of supply is to supply with decoding device with the video data of the compressed coding of codec processing unit.
17. messaging device according to claim 16, wherein control device forms codec processing unit, so that the sum that is included in I image in the codec processing unit of division or P image is than the image sets or the I image in formerly the codec processing unit or the total big number less than predetermined number of P image on the time that comprise in time formerly.
18. messaging device according to claim 3, wherein the video data of compressed coding comprises image sets, and
When reception comprises first image sets of I image or P image, the number of wherein said image deducts 3 values that obtain less than the group number from the group memory, go up control device detection time continuously in the structure of second image sets of first image sets, and, if the sum that is included in I image in first image sets and second image sets or P image is less than deducting 2 values that obtain by the group number from the group memory, control device merges first image sets and second image sets forming codec processing unit so, and the control apparatus for controlling of supply is to supply with decoding device with the video data of the compressed coding of codec processing unit.
19. messaging device according to claim 18, wherein control device form codec processing unit in case be included in I image in the codec processing unit of merging or the sum of P image than the I image in the image sets that comprises in time formerly and a total big number of P image less than predetermined number.
20. messaging device according to claim 3, wherein the video data of compressed coding comprises image sets, and
Control device detects the structure of first image sets and the structure of second image sets in first image sets continuous in time, if and the sum that is included in I image in first image sets and second image sets or P image is less than by deducting two times of 3 values that obtain from the group number of group memory, control device merges first image sets and second image sets so, divide the image sets merged then to form first codec processing unit and second codec processing unit, each codec processing unit comprises that number deducts the I image or the P image of 2 values that obtain less than the group number from the group memory, and control device control apparatus for controlling of supply is to supply with decoding device with the video data of the compressed coding of the video data of the compressed coding of first codec processing unit and second codec processing unit respectively.
21. messaging device according to claim 20, wherein control device forms first codec processing unit and second codec processing unit, so as the sum that is included in I image in first codec processing unit and second codec processing unit or P image than on the time formerly codec processing unit or on the time formerly the I image of image sets or a total big number of P image less than predetermined number.
22. an information processing method that is used for messaging device, the video data of described messaging device decoding compression-coded, this method may further comprise the steps:
The instruction of acquisition playback speed;
Determine to be included in the order of decoding processing of the image of the codec processing unit that is used for carrying out decoding processing, so as before the B image decoding I image and P image;
According to the playback speed instruction that is obtained, select the image that to export in the image from be included in codec processing unit;
In the image of control from be included in codec processing unit I image, P image and selected B image as the image that will export are supplied with decoder;
The image of decoder is supplied with in decoding; And
The output of the B image of the control image that selected conduct will be exported from decoded picture.
23. a messaging device that is used for the video data of decoding compression-coded comprises:
Be used to obtain the device of playback speed instruction;
Be used for determining the device of order of the decoding processing of image, wherein said image packets is contained in the codec processing unit that is used for carrying out decoding processing, so as before the B image decoding I image and P image;
Be used for according to the playback speed instruction that is obtained, the device of the image that selection will be exported among the image from be included in codec processing unit;
Be used for controlling from the image that is included in codec processing unit with I image and P image and the selected device that is supplied to decoder as the B image of the image that will export;
The image of decoder is supplied with in decoding; And
Control is selected as the output of wanting the B image of output image from the image of decoding.
24. a messaging device is used for the video data of decoding compression-coded, comprising:
Storage device is used to store the video data of compressed coding;
Read-out device is used for reading from storage device the video data of compressed coding;
Decoding device is used for the video data of decoding compression-coded;
Apparatus for controlling of supply, the video data that is used to control the compressed coding that will be read by read-out device is supplied to decoding device;
Control device is used to control the processing of being carried out by apparatus for controlling of supply and decoding device; And
The playback speed command device is used for playback speed is instructed apparatus for controlling of supply,
Wherein control device determines to be included in the order of decoding processing of the image of the codec processing unit that is used for the decoding processing carried out by decoding device, so as before the B image decoding I image and P image,
Control device is according to the playback speed instruction that is provided by the playback speed command device, and selecting in the image from be included in codec processing unit will be from the image of decoding device output,
I image, P image and selected conduct that control device control apparatus for controlling of supply is controlled in the image that will be included in the codec processing unit will be supplied to decoding device from the B image of the image of decoding device output, and
Control device control decoding device is supplied with the image of decoding device with decoding under the control of apparatus for controlling of supply, and control is as will be from the image of decoding device output and the output of selecteed B image.
25. an information processing method that is used for messaging device, the video data of described messaging device decoding compression-coded said method comprising the steps of:
Read the video data that is stored in the compressed coding in the memory cell;
The instruction of supply playback speed;
The playback speed that obtains to be supplied with instructs;
Determine to be included in the order of decoding processing of the image of the codec processing unit that is used for carrying out decoding processing, so as before the B image decoding I image and P image;
According to the playback speed instruction that is obtained, select the image that to export in the image from be included in codec processing unit;
The B image of the image that control will more be exported from I image, P image and selected conduct in the image the codec processing unit of the video data that is included in the compressed coding that reads is supplied to decoder; And
The image of decoder is supplied with in decoding; And
The output of the B image of the control image that selected conduct will be exported from the image of decoding.
26. a messaging device that is used for the video data of decoding compression-coded comprises:
Be used for reading the device of the video data of the compressed coding that is stored in memory cell;
Be used to supply with the device of playback speed instruction;
Be used to obtain the device of the playback speed instruction supplied with;
Be used for determining to be included in the order of decoding processing of the image of the codec processing unit that is used to carry out decoding processing, so as before the B image device of decoding I image and P image;
Be used for according to the playback speed instruction that is obtained, the device of the image that selection will be exported in the image from be included in codec processing unit;
The B image that is used for controlling the image that will will export from I image, P image and selected conduct the image of the codec processing unit of the video data that is included in the compressed coding that reads is supplied to the device of decoder;
The device of the image of supplying with decoder is used to decode; And
Be used for controlling the device of the output of the B image of selecting as output image more from the image of decoding.
27. an information processing method that is used for the video data of decoding compression-coded may further comprise the steps:
The video data of decoding compression-coded;
Control is supplied with decoder with the video data of compressed coding; And
The processing that control is carried out by supply controller and decoder,
Its middle controller is determined the order of the decoding processing that decoder is carried out, so that at the image that codec processing unit comprised that is used for the decoding processing that decoder carries out, and decoding I image and P image before the B image,
And
Controller is according to the playback speed instruction that the playback speed command unit provides, and selecting in the image from be included in codec processing unit will be from the image of decoder output;
Wherein said decoding step comprises:
Under the control of controller, among the image in being included in codec processing unit, with I image, P image and select to be supplied to the decoding processing final controlling element as the B image of the image of exporting from decoder, and
Under the control of controller, decoding is supplied with the image that controller is supplied with from image, and also output as will be from the image of decoder output and selecteed B image.
28. an information processing method that is used for the video data of decoding compression-coded may further comprise the steps:
Store the video data of compressed coding;
Read the video data of compressed coding from memory cell;
The video data of decoding compression-coded;
The video data of the compressed coding that control will be read by sensing element is supplied to decoder;
Control is by supplying with the processing that controller and decoder are carried out; And
Provide playback speed to instruct to controller,
Its middle controller determines to be included in the order of decoding processing of the image of the codec processing unit that is used for the decoding processing that decoder carries out, so as before the B image decoding I image and P image,
The playback speed instruction that controller provides according to the playback speed command unit, selecting among the image in being included in codec processing unit will be from the image of decoder output,
Controller control is supplied with controller with I image and the P image and selected as will from the B image of the image of decoder output be supplied to decoder of control among will the image from be included in codec processing unit, and
Controller control decoder decode is supplied with the image of decoder under the control of supplying with controller, and control is as will be from the image of decoder output and the output of selecteed B image.
CN200510121796A 2004-10-26 2005-10-26 Information processing apparatus and method Expired - Fee Related CN100592777C (en)

Applications Claiming Priority (3)

Application Number Priority Date Filing Date Title
JP2004311521 2004-10-26
JP2004311521 2004-10-26
JP2005242161 2005-08-24

Related Child Applications (1)

Application Number Title Priority Date Filing Date
CN2009102637371A Division CN101771843B (en) 2004-10-26 2005-10-26 Information processing apparatus and method

Publications (2)

Publication Number Publication Date
CN1893608A CN1893608A (en) 2007-01-10
CN100592777C true CN100592777C (en) 2010-02-24

Family

ID=37598056

Family Applications (1)

Application Number Title Priority Date Filing Date
CN200510121796A Expired - Fee Related CN100592777C (en) 2004-10-26 2005-10-26 Information processing apparatus and method

Country Status (1)

Country Link
CN (1) CN100592777C (en)

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102123280B (en) * 2010-01-11 2016-03-02 联发科技股份有限公司 Coding/decoding method and decoding device
US8570361B2 (en) 2010-01-11 2013-10-29 Mediatek Inc. Decoding method and decoding apparatus for using parallel processing scheme to decode pictures in different bitstreams after required decoded data derived from decoding preceding picture(s) is ready
CN103780942B (en) * 2013-12-23 2017-07-11 青岛海信电器股份有限公司 A kind of information configuring methods and device

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5739860A (en) * 1994-11-17 1998-04-14 Hitachi, Ltd. Method of and apparatus for decoding moving picture and outputting decoded moving picture in normal or frame-skipped reproducing operation mode
CN1451163A (en) * 2000-11-21 2003-10-22 皇家菲利浦电子有限公司 Arragnement for the reproduction of compressed data in a multi-speed mode

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5739860A (en) * 1994-11-17 1998-04-14 Hitachi, Ltd. Method of and apparatus for decoding moving picture and outputting decoded moving picture in normal or frame-skipped reproducing operation mode
CN1451163A (en) * 2000-11-21 2003-10-22 皇家菲利浦电子有限公司 Arragnement for the reproduction of compressed data in a multi-speed mode

Also Published As

Publication number Publication date
CN1893608A (en) 2007-01-10

Similar Documents

Publication Publication Date Title
CN101771843B (en) Information processing apparatus and method
US5621464A (en) Method of reordering a decoded video picture sequence
CN102595240B (en) Video signal/image-decoding system and corresponding video signal/image decoding method
US7515761B2 (en) Encoding device and method
JPH05183757A (en) Method and device for processing picture
CN100592777C (en) Information processing apparatus and method
RU2265879C2 (en) Device and method for extracting data from buffer and loading these into buffer
CN103986929A (en) Image processing device
CN102436363B (en) Method and device for automatically setting parameters of video card
JP2006157873A (en) Data processing system, reproduction apparatus and method, computer, program, and storage medium
JPH01278173A (en) Image reader
US6339669B1 (en) Picture-reproducing apparatus
CN107040784B (en) Video buffering processing method, system and device
EP1598827A1 (en) Video editing system and method
CN101090489B (en) Information processing apparatus and information processing method
US7729591B2 (en) Data processing apparatus, reproduction apparatus, data processing system, reproduction method, program, and storage medium
CN100534161C (en) Information recording apparatus
CN100499784C (en) Information processing apparatus and method
US7627232B2 (en) Reproduction apparatus, data processing system, reproduction method, program, and storage medium
US20040109505A1 (en) Image-processing method and image processor
JP2006330124A (en) Data processor
JP2006352451A (en) Image processor
KR100846406B1 (en) Video encoder transmitting extracted image outside during encoding
TW446888B (en) Animation display process method
JP2006157871A (en) Information processing apparatus and method, recording medium, and program

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C14 Grant of patent or utility model
GR01 Patent grant
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20100224

Termination date: 20151026

EXPY Termination of patent right or utility model