JP4315521B2 - Video compression encoding method and apparatus, and video compression encoding / decoding system - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a moving picture compression coding method and apparatus represented by the MPEG (Moving Picture Experts Group) system, and a moving picture compression coding / decoding system. More specifically, the present invention relates to a time for inverse 3: 2 pulldown processing. It is related to the technology of stamp making.
[0002]
[Prior art]
Inverse 3: 2 pull-down processing is processing for thinning out redundant fields from the NTSC signal subjected to 3: 2 pull-down processing. By the way, 3: 2 pull-down processing is a technique for converting moving image information equivalent to 1 second in 24 frames such as movie film into an NTSC television signal equivalent to 1 second in 30 frames.
[0003]
FIG. 9 is a schematic diagram showing the procedure of 3: 2 pull-down processing for converting a motion picture film frame into an NTSC television signal. FIG. 9 (a) shows four frames corresponding to 4/24 (1/6) second of the original film, which are F0, F1, F2, and F3 frames in order. When four frames F0, F1, F2, and F3 of this original film are directly converted into NTSC television signals, as shown in FIG. 9 (b), each frame is a top field (odd field) t0, t1, t2, t3. And bottom field (even field) b0, b1, b2, b3. That is, the frame F0 of the original film is in the fields t0 and b0 of the NTSC television signal, the frame F1 of the original film is in the fields t1 and b1 of the NTSC television signal, and the frame F2 of the original film is the field t2 of the NTSC television signal. In b2, the frame F3 of the original film is converted into fields t3 and b3 of the NTSC television signal, respectively.
[0004]
Then, the fields t0, b0, t1, b1, t2, b2, t3, b3 of the NTSC television signal obtained from the frames F0, F1, F2, F3 of the original film are shown in FIG. 9 (c). Thus, by using the fields t0 and b2 redundantly, 10 fields as an array of t0, b0, t0, b1, t1, b2, t2, b2, t3, b3, that is, f0, f1, f2, f3, Reconfigure to 5 frames of f4. Therefore, the frame f0 converted to the NTSC television signal is the fields t0 and b0, the frame f1 is the fields t0 and b1, the frame f2 is the fields t1 and b2, the frame f3 is the fields t2 and b2, and the frame f4 is the field f4 It consists of t3 and b3.
[0005]
As described above, motion picture information of a movie film corresponding to 1 second in 24 frames is converted into an NTSC television signal corresponding to 1 second in 30 frames (60 fields). By the way, when the television signal obtained in this way is compressed and encoded by the MPEG-2 system, the process of thinning out redundant fields included in two consecutive 10 fields, that is, inverse 3: 2 pull-down. It is necessary to perform processing.
[0006]
FIG. 10 is a block diagram showing a configuration example of a conventional moving picture compression encoding apparatus based on the MPEG-2 system. In FIG. 10, reference numerals 1 and 2 indicate first and second field memories. The first field memory 1 delays the input digital video signal by one field. Reference numeral 4 indicates a switch for switching whether or not the output from the first field memory is input to the second field memory 2. The output of the changeover switch 4, in other words, the first field memory 1 The second field memory 2 delays the output by one field. Accordingly, the output of the second field memory 2 is delayed by two fields from the input to the first field memory 1.
[0007]
Reference numeral 3 compares the input to the first field memory 1 with the output from the second field memory 2 (the input to the first field memory 1 is delayed by two fields), and is equivalent to one field. Is smaller than a predetermined value, more specifically, if the difference is substantially the same, the output of the changeover switch 4 is set to no signal, and otherwise, the output of the changeover switch 4 is set to the first field memory 1. The correlation detection circuit is switched so that the output is input to the second field memory 2.
[0008]
Reference numeral 5 is a video encoder that compresses the video signal, 100 is a PTS analysis circuit that analyzes the output data of the video encoder 5 and calculates a presentation time stamp (hereinafter referred to as PTS), and 101 is an output from the video encoder 5 The PTS analyzer circuit 100 packetizes the compressed data stored in the buffer memory 101 and the PTS analyzer circuit 100 calculates the compressed data stored in the buffer memory 101 while the PTS analyzer circuit 100 analyzes the PTS. This is a PTS addition circuit that adds a PTS and outputs it.
[0009]
The first field memory 1, the second field memory 2, the correlation detection circuit 3, and the changeover switch 4 constitute an inverse 3: 2 pull-down processing unit 50. The inverse 3: 2 pull-down processing unit 50 compares each field of the input video data with the field two fields before, and when the difference between the two is smaller than a predetermined value, that is, when the correlation is large, the same field is used. By determining that there is a switch and switching the changeover switch 4 to the non-signal side, reverse 3: 2 pull-down processing for removing redundant fields is performed.
[0010]
The output signal of the changeover switch 4 is input to the video encoder 5, and when the changeover switch 4 is switched to the non-signal side, the video encoder 5 repeats the first field of the frame to be input again at the time of decoding. An MPEG-2 video elementary compression encoding process is performed after enabling an RFR (Repeat First Field) flag for output.
[0011]
In MPEG-2 video elementary compression coding, compression coding processing using correlation of moving images is performed, so the frame arrangement after encoding (hereinafter referred to as picture arrangement) is different from the picture arrangement of the original signal. In order. A typical example of this state is shown in the schematic diagram of FIG. 11 when the period in which an I or P picture appears is three. Note that an I picture (Intra coded picture) is called an intra-frame coded picture, and is coded independently in only one frame. A P picture (Predictive coded picture) is called a forward predictive coded picture, and is encoded from an I picture by forward prediction. In addition to this, there is a B picture (Bidirectionally predictive coded picture) called a bidirectional predictive coded picture that is predictively coded using the preceding and following I and P pictures as reference pictures.
[0012]
In the 3: 2 pull-down processed NTSC signal shown in FIG. 11 (a), the inverse 3: 2 pull-down processing shown in FIG. 11 (b) and the playback image shown in FIG. 11 (f), t0, t1,... represent the top fields of frames 0, 1,..., and b0, b1. In the video encoding process shown in (c) and the decoding process shown in (e), for example, B0, I2, and P5 are picture types (B, I, P) and frame numbers (0, 2 and 6) respectively.
[0013]
One field (t0, b2, t4, b6, t8, b10) is redundantly added with a period of 4 fields to the 3: 2 pull-down processed NTSC signal shown in FIG. 11 (a). A signal obtained by digitally converting this signal is the bidet data input shown in FIG. 10, and redundant redundant fields are removed by the inverse 3: 2 pull-down processing unit 50, and video encoding processing is performed by the video encoder 5. It is.
[0014]
In the video encoding process by the video encoder 5, since the B picture is predictively encoded using the preceding and subsequent I and P pictures as reference images, the I and P pictures to be referred to are encoded first, and then the B pictures are encoded. Encoded. For this reason, although the details will be described later, the order of the I, P picture and B picture is switched.
[0015]
The PTS analysis circuit 100 sequentially analyzes the encoded data output from the video encoder 5 to extract the picture type (I, P or B), the RFF flag, and the frame period, and calculates the PTS based on these.
[0016]
However, for I and P pictures, PTS cannot be calculated unless the state of the RFF flag of the B picture behind it is known. For example, in order to calculate the PTS of the I2 picture shown in FIG. 11 (c), the RFF flag of the B0 and B1 pictures preceding it is confirmed. If the RFF flag is valid “1”, one field The PTS of minutes must be added. In the example shown in FIG. 11, the PTS of the I2 picture is 5 fields in terms of the number of fields because the RFF flag of the B0 picture is valid as shown in FIG. 11 (d).
[0017]
Thus, since the PTS of the I and P pictures is not determined unless the information of the B picture immediately before the next I or P picture is known, the buffer memory 101 that holds and delays the data until then is required. It becomes.
[0018]
Therefore, the buffer memory 101 is controlled so that data is read when the PTS analysis circuit 100 determines the PTS. The data read from the buffer memory 101 is packetized by the PTS adding circuit 9, and the PTS calculated by the PTS analyzing circuit 100 is added and output.
[0019]
In addition, as an example of the case where a moving image compression encoding apparatus is configured in combination with a general-purpose computer such as a computer, up to the video encoder 5 for creating an elementary stream, that is, the inverse 3: 2 pull-down processing unit 50 and the video encoder 5 are used. Can be configured by dedicated hardware, and the subsequent functions of the PTS analysis circuit 100 and the PTS additional circuit 9 can be performed by software processing.
[0020]
However, even in such a case, the same processing procedure is necessary. While encoding results are sequentially analyzed by software corresponding to the PTS analysis circuit 100, the encoded data is packetized by software corresponding to the PTS additional circuit 9 and PTS. A video PES (Packetized Elementary Stream) is output by adding a PTS calculated by software corresponding to the analysis circuit 100.
[0021]
After that, an audio PES encoded with an audio encoder (not shown) and added with PTS, an SCR (System Clock Reference) serving as a reference reference value for a time stamp, and the video PES described above are connected with a multiplexer (not shown). A system stream is created by multiplexing. The system stream created in this manner is reproduced by the decoding device while synchronizing video and audio using video and audio PTS and SCR.
[0022]
[Problems to be solved by the invention]
Since the conventional moving image compression coding apparatus is configured as described above, when determining the time stamp information of certain picture data, information on the picture data output after the picture data is necessary. For this reason, a buffer memory for holding all data up to the picture data to be output later is required. In addition, since the capacity of the buffer memory is set assuming the maximum encoding rate and the maximum I / P picture interval, there is a problem that a large capacity is required.
[0023]
On the other hand, since the encoding rate actually used and the actual I / P picture interval are usually less than the assumed maximum value, there is a problem that a part of the memory becomes redundant and cannot be used effectively. .
[0024]
In addition, a general-purpose computer such as a personal computer is connected to a board composed of dedicated 3: 2 pull-down processing unit 50 and video encoder 5, and the subsequent functions of PTS analysis circuit 100 and PTS additional circuit 9 are software. Even in the case of adopting a configuration performed by processing, a memory capacity corresponding to the buffer memory 101 is required for the internal memory of the computer.
[0025]
The present invention has been made in view of such circumstances, and provides a moving image compression encoding method and apparatus, and a moving image compression encoding / decoding system that do not require a large-capacity and redundant buffer memory. With the goal.
[0026]
[Means for Solving the Problems]
According to a first aspect of the present invention, there is provided a moving image compression encoding method, wherein each frame of an original moving image is converted into a moving image in which one frame is composed of two fields, and a predetermined number of consecutive frames of the moving image are converted into one. Detecting a duplicate field in an image signal obtained by adjusting a time axis in advance by duplicating a specific field in each unit frame as a unit, and deleting one redundant field; A step of outputting field deletion information indicating that the image signal has been deleted, a step of compressing and encoding the image signal after the redundant field is deleted, and the number of fields at the time when the image signal is compressed and encoded in units of frames. Cumulative field count plus field delete count Total And outputting picture type information identifying the type of image data defining the order of arrangement with respect to other frames after the frame is compression encoded and counted Cumulative field count And picture type information are stored for each picture type information based on the output picture type information. Cumulative field count Corresponding to compression-encoded picture type information Cumulative field count Oldest of Cumulative field number Selecting sequentially, Counting the time information by converting the selected cumulative field number to the reference frequency clock number, deleting the selected cumulative field number, and counting Of adding the time information thus made to a frame unit of a compression-encoded image signal And It is characterized by including.
[0027]
In such a moving image compression encoding method according to the first aspect of the present invention, the time information counted when the image data is compression encoded and the picture type information after the compression encoding are held, and the compression encoding process is performed. The time information corresponding to the same picture type information as the picture type information of the frame of the image data thus selected is selected, and the time information is determined. Therefore, in order to calculate time information of a frame of certain image data, it is not necessary to perform a complicated process such as calculating time information after detecting an RFF flag which is field deletion information of a subsequent frame.
[0028]
The moving image compression encoding method according to claim 2 of the present invention is: By converting each frame of the original moving image into a moving image in which one frame is composed of two fields, and overlapping a specific field in each unit frame with a predetermined number of consecutive frames of the moving image as one unit. A step of detecting a duplicate field in an image signal obtained by adjusting a time axis in advance and deleting one redundant field; a step of outputting field deletion information indicating that a field has been deleted; and a redundant field Compressing and encoding the image signal after the image is deleted, generating a clock as a reference for a time stamp indicating time information, and counting the generated clocks to compress the image data in units of frames. The time stamp at the time of conversion is calculated, and other frames after the frame is compressed and encoded A step of outputting picture type information for specifying the type of image data defining the arrangement order to be stored, and storing the calculated time stamp and picture type information, corresponding to the picture type information of the image data after compression coding Sequentially selecting from the oldest time stamp among the same picture type information as the picture type information of the image data after compression encoding among the time stamps in which the time stamps to be stored are stored, and after the compression encoding Adding a selected time stamp to the data of each frame of It is characterized by that.
[0029]
A moving image compression encoding apparatus according to claim 3 of the present invention is provided. By converting each frame of the original moving image into a moving image in which one frame is composed of two fields, and overlapping a specific field in each unit frame with a predetermined number of consecutive frames of the moving image as one unit. A duplicate field in an image signal obtained by adjusting the time axis in advance is detected, one redundant field is deleted, field deletion information indicating that the field has been deleted, and an image after the redundant field is deleted Field decimation means for outputting a signal, and an image signal and field deletion information output from the field decimation means, and a compression code for compressing and encoding the image signal and outputting the number of compressed fields and field deletion information And other frames of each frame after compression encoding of the image signal input to the compression encoding means. A field number calculating unit that outputs picture type information for specifying a type of image data that defines an arrangement order for a frame, and calculates a cumulative number of fields at the time of input to the field thinning unit; and the field number calculating unit The compression encoding means of the cumulative field number stored in the stored field number and the picture type information calculated from the above, and the cumulative field number corresponding to the picture type information output from the compression encoding means A field number selection means for sequentially selecting from the smallest accumulated field number in the same picture type information as the picture type information output by; a time stamp conversion means for converting the accumulated field number into a time stamp representing time; and a compression code Time stamp conversion means for each frame data after conversion And a time information adding means for adding the converted time stamp It is characterized by that.
[0030]
Such claims 3 In the moving image compression coding apparatus described in the above, the cumulative field number counted at the time of input to the compression coding means and the picture type information after the compression coding are held and output from the compression coding processing means. The time information is calculated by a simple conversion of selecting the number of accumulated fields corresponding to the same picture type information as the picture type information. Therefore, in order to calculate time information of a frame of certain image data, it is not necessary to perform a complicated process such as calculating time information after detecting an RFF flag which is field deletion information of a subsequent frame.
[0031]
According to a fourth aspect of the present invention, there is provided a moving picture compression encoding apparatus that converts each frame of an original moving picture into a moving picture having one frame consisting of two fields, and a predetermined number of consecutive frames of the moving picture are converted into one. By duplicating a specific field in each unit frame as a unit, a duplicate field in the image signal obtained by adjusting the time axis in advance is detected, and one redundant field is deleted, and the field is deleted. Field deletion information indicating that this is the case and a field thinning means for outputting an image signal after redundant fields are deleted, and an image signal and field deletion information output from the field thinning means are input, and the image signal is compressed and encoded And a compression encoding means for outputting the number of fields subjected to compression processing and field deletion information, and a time star representing time information. A clock generation unit that generates a clock serving as a reference for the image; and a time stamp at which the image data is input to the compression encoding unit in units of frames by counting the clocks generated by the clock generation unit A time stamp calculating means for outputting picture type information for specifying a type of image data defining an arrangement order with respect to another frame after the frame is compression-encoded, and a time stamp calculated by the time stamp calculating means And the picture type information, and a time stamp corresponding to the picture type information output from the compression encoding means is stored. Among the time stamps of the same picture type information as the picture type information output from the compression encoding means A time stamp selecting means; and a time stamp adding means for adding the time stamp selected by the time stamp selecting means to the data of each frame after compression encoding.
[0033]
Such claims 4 In the moving image compression coding apparatus described in the above, the time stamp counted at the time of input to the compression coding means and the picture type information after compression coding are held and output from the compression coding processing means. The time stamp is determined by selecting the time stamp corresponding to the same picture type information as the selected picture type information. Therefore, in order to calculate the time stamp of a frame of a certain image data, it is not necessary to perform a complicated process such as calculating a time stamp after detecting an RFF flag which is field deletion information of a subsequent frame.
[0034]
Claims of the invention 5 The moving image compression encoding / decoding system described in 1) converts each frame of the original moving image into a moving image in which one frame is composed of two fields, and each of the predetermined number of consecutive frames of the moving image as one unit. By duplicating a specific field in one unit frame, a duplicate field in an image signal obtained by adjusting the time axis in advance is detected, and one redundant field is deleted, indicating that the field has been deleted. Field decimation means for outputting an image signal after deletion of field deletion information and redundant fields, and an image signal and field deletion information output from the field decimation means are input, and the image signal is independently transmitted in frame units. Using the I-picture that is an intra-frame coded picture to be compression-coded and the correlation characteristics between the I-picture and the frame A P picture that is a forward-predictive encoded picture that is compression-encoded and a B picture that is a bi-predictive encoded picture that is compression-encoded using the correlation characteristics between the I and P pictures. Compression encoding means for performing compression encoding and outputting the number of fields subjected to compression processing and field deletion information, and obtaining a repetition cycle of an I picture or P picture after encoding of an image signal input to the compression encoding means And an information extracting means for extracting picture type information for specifying the type of image data defining the arrangement order with respect to other frames after the frame is compression-encoded and field deletion information, and the information extracting means The repetition period and picture type information of the acquired or extracted I picture or P picture Therefore, when the picture type is an I picture or P picture, the number of fields corresponding to the repetition period and the next B of the previous I picture or P picture are added to the cumulative number of fields of the previous I picture or P picture. The number of field deletions from the picture to the picture is added to obtain the cumulative field number of the picture. If the picture type is a B picture and the previous picture type is an I picture or P picture, then two The cumulative field number of the picture is obtained by adding “2” to the cumulative number of fields of the previous I picture or P picture and the B picture field deletion count from the B picture next to the previous I picture or P picture to the picture. The picture type is B picture and the previous picture type is B If a puncture is a cumulative number field of the previous B the picture by adding the a field deletion count of the picture "2" of the number of cumulative field picture Field number predicting means, time stamp converting means for converting the cumulative number of fields into image time stamps, and adding image time stamps to the data after being compression encoded by the compression encoding means, Image time stamp adding means for outputting, audio time stamp adding means for adding audio time stamp to audio data and outputting audio packet data, image packet data output from the image time stamp adding means, and audio time Multiplexing means for multiplexing and outputting voice packet data output from the stamp adding means and a reference time stamp, and data for separating the data multiplexed by the multiplexing means into image packet data and voice packet data Separation means and voice time data from voice packet data Audio synchronizing processing means for extracting audio data and performing audio synchronizing processing, audio decoding means for decoding and outputting audio packet data, and a time stamp of an image extracted from the image packet data and a reference time stamp Time-axis filter means for performing time-axis filter processing on the difference, image synchronization processing means for performing image synchronization processing with a value output from the time-axis filter means, and an image for decoding and outputting image packet data And a decoding means.
[0035]
Such claims 5 In the moving image compression encoding / decoding system described in 1), time stamps of I and P pictures are calculated at the time of encoding. Therefore, it is not necessary to perform complicated processing such as calculating the time stamp after detecting the RFF flag of the B picture appearing thereafter. In addition, a stable synchronization process is performed at the time of decoding.
[0037]
Claims of the invention 6 The video compression encoding / decoding system described in claim 1 5 In the moving image compression encoding / decoding system, the time axis filter unit adds a difference between a time stamp of an image extracted from image packet data and a reference time stamp for a predetermined number of pictures, and adds the result of the addition to the time stamp. Filtering in the time axis is performed by using the quotient divided by the predetermined number as a difference, and the predetermined number is a common multiple of the repeated picture period of the field deletion information and the repeated picture period of the I picture or P picture. It is characterized by.
[0038]
Claims of the invention 7 The video compression encoding / decoding system described in claim 1 5 In the moving image compression encoding / decoding system, the time axis filter unit adds a difference between a time stamp of an image extracted from image packet data and a reference time stamp for a predetermined number of pictures, and adds the result of the addition to the time stamp. Filtering in the time axis is performed by using a quotient divided by a predetermined number as a difference, and the predetermined number is a least common multiple of a repeated picture period of field deletion information and a repeated picture period of an I picture or a P picture. It is characterized by that.
[0039]
Such claims 5, 6 and 7 In the moving image compression encoding / decoding system according to the present invention described in 1), a synchronization matching process with a quick response is performed.
[0040]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, a moving picture compression coding apparatus and a moving picture compression coding / decoding system according to the present invention will be described in detail with reference to the drawings showing the respective embodiments. In addition, in each figure referred in the following description and each figure referred in description of a prior art example, the part shown with the same code | symbol shows the same or equivalent part.
[0041]
Embodiment 1 FIG.
FIG. 1 is a block diagram showing a configuration example of a moving image compression coding apparatus according to the present invention. In FIG. 1, reference numerals 1 and 2 indicate first and second field memories. The first field memory 1 delays the input digital video signal by one field. Reference numeral 4 indicates a switch for switching whether or not the output from the first field memory is input to the second field memory 2. The output of the changeover switch 4, in other words, the first field memory 1 The second field memory 2 delays the output by one field. Therefore, the output of the second field memory 2 is delayed by two fields from the input to the first field memory 1.
[0042]
Reference numeral 3 compares the input to the first field memory 1 with the output from the second field memory 2 (the input to the first field memory 1 is delayed by two fields), and is equivalent to one field. Is smaller than a predetermined value, more specifically, if the difference is substantially the same, the output of the changeover switch 4 is set to no signal, and otherwise, the output of the changeover switch 4 is set to the first field memory 1. The correlation detection circuit is switched so that the output is input to the second field memory 2.
[0043]
Reference numeral 5 is a video encoder for compressing the video signal, 6 is a field number counting circuit for counting the number of fields of the compressed data output from the video encoder 5, 7 is a field number replacement circuit, and 8 is field 9 shows a PTS conversion circuit that converts the processing result of the number change circuit 7 into PTS, and 9 shows a PTS addition circuit that adds the PTS converted by the PTS conversion circuit 8 to the compressed data output from the video decoder 5 and outputs it. .
[0044]
Next, the operation of the moving picture compression encoding apparatus according to the present invention shown in FIG. 1 will be described. The first field memory 1, the second field memory 2, the correlation detection circuit 3, and the changeover switch 4 constitute an inverse 3: 2 pull-down processing unit 50. The inverse 3: 2 pull-down processing unit 50 is the same in configuration and operation as before, and compares each field of the input video data with the field two fields before, and if the difference between the two is smaller than a predetermined value, When the correlation is large, it is determined that they are the same field, and the selector switch 4 is switched to the non-signal side, thereby performing an inverse 3: 2 pull-down process for removing redundant fields.
[0045]
The output signal of the changeover switch 4 is input to the video encoder 5, and when the changeover switch 4 is switched to the non-signal side, the video encoder 5 repeats the first field of the frame to be input again at the time of decoding. RF, which is field deletion information for output F ( The MPEG-2 video elementary compression / encoding process is performed with the Repeat First Field) flag enabled, and the resulting video elementary stream is output to the PTS adding circuit 9 while being packetized.
[0046]
Further, each time video data is input in units of frames, the video encoder 5 outputs the picture type (I, P, B) after the encoding of the input data and the presence / absence of the RFF flag to the field number counting circuit 6. The picture type is also output to the field number replacement circuit 7.
[0047]
The field number counting circuit 6 executes a processing sequence as shown in the flowchart of FIG. First, the field number counting circuit 6 initializes the cumulative field number F (step S11), and then every time video data is input to the video encoder 5 in units of frames (step S12), the picture type after the input data is encoded. (I, P, B) and the presence / absence of the RFF flag are obtained from the video encoder 5 (step S13), and the accumulated field number and picture type of the current picture already counted are output to the field number replacement circuit 7 ( Step S13).
[0048]
When the RFF flag is valid (“YES” in step S14), the field number counting circuit 6 counts the number of fields of the picture as “3” (step S15), and when it is invalid (step S15). "NO" in S14), count as "2" (step S16), and by adding to the cumulative field number of the current picture, it is held as the cumulative field number of the next picture to be acquired (step S17).
[0049]
The field number replacement circuit 7 includes a register for storing picture types and cumulative field numbers corresponding to the maximum period of I and P pictures and the number of pictures corresponding to the maximum delay time of I and P pictures in the video encoder 5. Has been. For example, when the maximum period of I and P pictures is 3, and the maximum encoding delay time of I and P pictures is 0.4 frames, the added value “3.4” of both is rounded up and the registers for 4 pictures are fielded. The number change circuit 7 is configured. In these registers, the cumulative field number and picture type of the picture output from the field number counting circuit 6 are sequentially stored.
[0050]
The field number counting circuit 6 executes a processing sequence as shown in the flowchart of FIG. Whenever the video data is input to the video encoder 5 in units of frames (step S21), the field number changing circuit 7 acquires the picture type output from the video encoder 5 (step S22), and the timing at which the picture is output. The number of accumulated fields corresponding to the oldest picture of the same picture type stored in the register is output (step S23). Note that the register that has output the cumulative number of fields is cleared (step S24), and preparations for storing new information are made.
[0051]
The PTS conversion circuit 8 is output from the field number replacement circuit 7. Accumulation When the number of fields is input, PTS is calculated from the following equation (1) and output to the PTS adding circuit 9. PTS is 90kHz count, video is NTSC signal, frame frequency is 29.9 7H z.
[0052]
PTS = Cumulative field number x 1 / 29.97 x 1/2 x 90000 (1)
[0053]
The PTS adding circuit 9 outputs a video PES (Packetized Elementary Stream) by adding the PTS output from the PTS conversion circuit 8 while packetizing the video elementary stream output from the video encoder 5.
[0054]
Embodiment 2. FIG.
By the way, in the first embodiment described above, as shown in FIG. 1, the moving picture compression coding apparatus according to the present invention calculates the PTS, as shown in FIG. 1, the field number counting circuit 6, the field number replacement circuit 7, and the PTS conversion. Although the circuit 8 is provided, it is of course possible to adopt other configurations. A moving image compression coding apparatus according to the present invention having such a configuration will be described below as a second embodiment.
[0055]
FIG. 4 is a block diagram showing a configuration example of a moving image compression encoding apparatus according to the present invention having a configuration different from that of the first embodiment as a configuration for calculating the PTS. In the first embodiment, the PTS is calculated by the field number counting circuit 6, the field number switching circuit 7, and the PTS conversion circuit 8. In the second embodiment, the 90 kHz frequency divider is used. 10, the PTS count circuit 11, and the PTS replacement circuit 12 are configured to calculate PTS.
[0056]
The 90 kHz frequency divider 10 divides the 27 MHz clock, which is the system reference clock, to generate a 90 kHz clock signal that serves as a PTS reference, and inputs it to the PTS count circuit 11.
[0057]
The PTS count circuit 11 starts counting the clock signal input from the 90 kHz frequency divider 10 using the encoding start information output from the video encoder 5 as a trigger. Then, the PTS count circuit 11 acquires the picture type after the input data is compression-coded for each frame at the time when the video data is input to the video encoder 5, and the count value at the time of acquisition is used as the PTS of the picture. To the PTS replacement circuit 12 together with the picture type.
[0058]
The PTS replacement circuit 12 has a register for storing the picture type of the picture corresponding to the sum of the maximum period of the I and P pictures and the maximum delay time of the I and P pictures in the video encoder 5 and the number of accumulated fields. It is configured. For example, if the maximum period of I and P pictures is 3, and the maximum encoding delay time of I and P pictures is 0.4 frames, the registers for 4 pictures obtained by rounding up the addition value “3.4” of both are replaced with PTS. The circuit 12 is configured. In each of these registers, the PTS and the picture type of each picture output from the PTS count circuit 11 are sequentially stored.
[0059]
When the PTS replacement circuit l2 acquires the picture type output from the video encoder 5, the PTS corresponding to the oldest picture of the same picture type stored in the register at the timing when the corresponding picture is output. Is output. Note that after outputting the PTS, the PTS replacement circuit 12 clears the corresponding register and prepares to store new information.
[0060]
The PTS adding circuit 9 outputs a video PES (Packetized Elementary Stream) by adding the PTS output from the PTS replacement circuit l2 while packetizing the video elementary stream output from the video encoder 5.
[0061]
Embodiment 3 FIG.
By the way, in the moving picture compression encoding apparatus according to the present invention of Embodiments 1 and 2 described above, the PTS is calculated by obtaining the cumulative field number or PTS at the time of input to the video encoder 5. It is also possible to adopt another configuration, for example, a configuration in which the data output from the video encoder 5 is analyzed and the PTS is simply predicted. A moving image compression encoding / decoding system according to the present invention having such a configuration will be described in detail below as a third embodiment.
[0062]
FIG. 5 is a block diagram showing a configuration example of a moving image compression encoding / decoding system according to the present invention having a configuration for simply predicting PTS by analyzing data output from the video encoder 5. The reverse 3: 2 pull-down processing unit 50, the video encoder 5 and the PTS adding circuit 9 are provided in the same manner as in the first and second embodiments shown in FIGS. The provision of the circuit 8 is the same as that of the first embodiment shown in FIG.
[0063]
In FIG. 5, reference numeral 13 is an information extraction circuit for extracting information from data encoded by the video encoder 5, and 14 is a field number simple prediction circuit for predicting the cumulative number of fields based on the information extracted by the information extraction circuit 13. The number of fields predicted by the simple number-of-fields prediction circuit 14 is given to the PTS conversion circuit 8 and converted to PTS, and the result is output to the PTS addition circuit 9.
[0064]
Reference numeral 15 is an audio encoder that compresses and encodes audio data, 16 is a PTS addition circuit that adds PTS by converting the audio data compression encoded by the audio encoder 15 into a packet (PES), and 26 is a time stamp. SCR creation circuit for generating SCR (System Clock Reference) as reference reference value, 17 multiplexes video PES data and audio PES data created by PTS additional circuits 9 and 16 and SCR created by SCR creation circuit 26 Each of the multiplexing circuits is shown.
[0065]
The multiplexed compressed data output from the multiplexing circuit 17 is recorded by a recording device 18 on a recording medium such as a magnetic disk.
[0066]
On the other hand, reference numeral 19 is a separation circuit that separates data read from the recording medium by the storage device 18 into video PES data, audio PES data, and SCR, and 20 is an audio that decodes the audio PES data separated by the separation circuit 19. Decoder, 21 is a video decoder for decoding video PES data separated by separation circuit 19, 22 is a reference time generation circuit for generating a reference time from SCR separated by separation circuit 19, and 23 is an audio encoder 20 is an audio PTS comparison circuit that compares the audio PTS encoded by the reference time generation circuit 22 with the reference time generated by the reference time generation circuit 22 and outputs a difference, and 24 is the video PTS encoded by the video encoder 21 and the reference time generation circuit 22. The video PTS comparison circuit that outputs the difference by comparing with the generated reference time, and 25 is the difference value of the video PTS output by the video PTS comparison circuit 24. A PTS differential filter to synchronous control information between the video and audio to filter on during shaft.
[0067]
The information extraction circuit 13 acquires the period M of the I and P pictures from the video encoder 5, extracts the picture type and RFF flag from the data encoded by the video encoder 5, and outputs them to the field number simple prediction circuit 14. .
[0068]
The field number simple prediction circuit 14 predicts the cumulative field number of the current picture in a processing sequence as shown in the flowchart of FIG. An outline of the sequence shown in FIG. 6 is as follows. First, the number-of-fields simple prediction circuit 14 performs compression encoding processing based on the period of I and P pictures and the picture type of the current picture output from the video encoder 5. The cumulative number of fields in which the rearrangement of the picture order occurring before and after is predicted is calculated. Furthermore, when the RFF flag of the current picture is valid, the field number simple prediction circuit 14 adds the result of adding “1” to the cumulative field number of the current picture that has already been predicted and calculated. And Specifically:
[0069]
For example, when the period M of I and P pictures is 3, the number of accumulated fields predicted by the sequence shown in FIG. 6 is shown in the schematic diagrams of FIGS. 7 and 8 are originally a diagram in which the right end of FIG. 7 and the left end of FIG. 8 are connected.
[0070]
7 and 8, the NTSC signal input to the moving image compression coding apparatus is obtained by performing 3: 2 pulldown conversion from the film material, and includes the first field memory 1, the second field memory 2, and the correlation detection. When reverse 3: 2 pull-down processing is performed in the reverse 3: 2 pull-down processing unit 50 constituted by the circuit 3 and the changeover switch 4, the RFF flag becomes valid every other frame and is input to the video encoder 5 ( Line 2). In the video encoder 5, the input image is encoded in the order of picture types of B, B, I, B, B, P... (Line 1). In this case, the correct number of accumulated fields is a value as shown in row 3.
[0071]
In the encoding process in the video encoder 5, since the B picture is predictively encoded using the preceding and following I and P pictures as reference images, the order of the I, P and B pictures is switched (line 4), the RFF flag and the correct accumulation The number of fields is also shown in line 5 and line 6 accordingly.
[0072]
The accumulated field number predicted by the sequence shown in FIG. As shown in FIG. 6, the cumulative field number IP of the previous I or P picture (line 7), the cumulative field number IP1 (line 8) of the previous I picture or P picture, and the I or P picture The effective number BR (line 9) of the RFF flag of the B picture sandwiched between the two, and the effective number BR1 (line 10) of the RFF flag of the previous B picture sandwiched between the I or P picture, From the period M of the I or P picture, the cumulative field number F (row 11) of each picture is predicted.
[0073]
In the prediction sequence of FIG. 6, first, the above-described variables are initially set (step S31). Specifically, the predicted field number F is “0”, the previous I and P picture cumulative field number IP is “−2”, and the previous I and P picture cumulative field number IP1 is “−8”. The effective frequency BR of the RFF flag of the B picture sandwiched between the I and P pictures is “0”, the effective frequency BR1 of the RFF flag of the previous consecutive B picture is “0”, and the cycle M of the I and P pictures is Initially set to “3”.
[0074]
Next, the picture type and the RFF flag are acquired from the data output of the video encoder 5 (step S32). If the acquired picture type is an I or P picture (“YES” in step S33), the current picture and the current picture are added to the previous I or P picture cumulative field number IP by the following equation (2). By adding the number of fields (2 × M) between the previous I or P pictures and adding the effective RFF number (BR) between them, a predicted value of the cumulative number of fields F is obtained (step S34).
[0075]
F = IP + 2 x M + BR (2)
[0076]
Next, only when the RFF flag of the current picture is valid (“YES” in step S35), “1” is added to the cumulative field number F (step S36) and output as the predicted cumulative field number P of the current picture. (Step S38). At this time, IP1 is replaced with IP (line 8), IP is replaced with F (line 7), BR1 is replaced with BR (line 10), and BR is replaced with “0” (line 9) (step S37).
[0077]
Next, when the picture type acquired from the data output of the video encoder 5 is a B picture ("NO" in step S33) and is acquired immediately after an I or P picture ("YES" in step S41). ), “2” is added as the number of fields before compression encoding between the current picture and the previous I or P picture to the cumulative field number IP of the previous I or P picture by the following equation (3): Further, the RFF effective count (BR1) of the B picture between the previous I or P picture and the previous I or P picture is added (step S42).
[0078]
F = IP1 + 2 + BR1 (3)
[0079]
Only when the RFF flag of the current picture is valid (“YES” in step S43), “1” is added to F and output as the cumulative field number F of the current picture, and when the RFF flag is valid “1” is added to BR (step S44).
[0080]
Next, when the picture type obtained from the data output of the video encoder 5 is a B picture ("NO" in step S33), and the previous picture is a B picture ("NO" in step S41). Then, “2” is added to the cumulative field number F (step S45).
[0081]
Only when the RFF flag of the current picture is valid (“YES” in step S43), “1” is added to F and output as the cumulative field number F of the current picture, and when the RFF flag is valid “1” is added to BR (step S44).
[0082]
The cumulative field number (line 11) predicted by the simple field number prediction circuit 14 in this way is a difference of ± 1 field compared to the correct cumulative field number (line l2). The reason is that the validity / invalidity of the RFF flag is reflected in the number of accumulated fields of the picture, and ± 1 in the least common multiple of 6 frames of 3 repetition cycles of I and P pictures and 2 repetition cycles of the RFF flag. The field difference is repeated. Then, by applying a filter on the time axis by the PTS difference filter 25 at the time of decoding, synchronization can be performed with high accuracy having no practical problem of ± 1 field or less.
[0083]
In addition, when such a prediction method is adopted, image data that has been converted to an NTSC signal without being subjected to 3: 2 pull-down processing from film material, that is, image data that is an NTSC signal obtained directly from the original signal is input. In such a case, the inverse 3: 2 pull-down process in the inverse 3: 2 pull-down processing unit 50 is not performed, so the RFF flag remains invalid. For this reason, the difference between the cumulative field number predicted from the period of I and P pictures and the picture type and the correct cumulative field number is “0”, and no unnecessary offset occurs.
[0084]
Thus, the predicted cumulative field number of the current picture output from the field number simple prediction circuit 14 is converted from the field number to PTS by the PTS conversion circuit 8 according to the equation (1) and output to the PTS adding circuit 9.
[0085]
The PTS addition circuit 9 outputs a video PES (Paeketized Elementary Stream) by adding the PTS output from the PTS conversion circuit 8 while packetizing the video elementary stream from the video encoder 5.
[0086]
On the other hand, the audio data is compressed and encoded by the audio encoder 15, and the PTS adding circuit 16 adds the PTS of the audio data to form a packet (PES).
[0087]
Also, an SCR (System Clock Reference) serving as a reference reference value for the time stamp is generated by the SCR generating circuit 26 and output to the multiplexing circuit 17.
[0088]
The video PES data, the audio PES data, and the SCR are multiplexed by the multiplexing circuit 17 into one data, and are recorded on a recording medium such as a magneto-optical disk by the recording device 18.
[0089]
The data read from the recording medium by the recording device 18 is separated into video PES data, audio PES data, and SCR by the separation circuit 19, and the audio PES data is decoded by the audio decoder 20, and the video PES data is decoded by the video decoder 21. The SCR is input to the reference time generation circuit 22 after being decoded.
[0090]
When the video PES data is decoded by the video decoder 21, the I and P pictures become reference pictures of B pictures. The field number difference (line 14) is also changed accordingly.
[0091]
The reference time generation circuit 22 refers to the SCR input from the separation circuit 19 and generates a reference time. Since the reference time corresponds to the correct cumulative field number (line 3, line 6) shown in FIGS. 7 and 8, the cumulative field number difference of line 14 is the predicted cumulative field number (line 1) and the reference time field. It is equal to the difference from the number conversion value.
[0092]
The audio PTS comparison circuit 23 compares the audio PTS extracted by the audio decoder 20 with the reference time, and outputs the difference value. The audio decoder 20 performs synchronization adjustment processing so that the difference value is “0”.
[0093]
Similarly, the video PTS comparison circuit 24 compares the video PTS extracted by the video decoder 21 with the reference time, and outputs the difference value. The PTS difference filter 25 holds the difference value output from the video PTS comparison circuit 24 for 6 pictures, and calculates the average value for every 6 pictures. The calculation result is shown in line 15 of FIGS. The initial value is “0”.
[0094]
This calculation result is divided by the time corresponding to one field, and the video decoder 21 performs synchronization control of the number of fields corresponding to the quotient. Further, the remainder of the above division is input to the audio PTS comparison circuit 23 and added to the difference value between the audio PTS and the reference time. The addition result is input to the audio decoder 20, and synchronization processing is performed so that the difference value becomes “0”. That is, the fraction of one field or less synchronizes audio with video by controlling audio synchronization processing.
[0095]
By performing such synchronization processing, video and audio can be synchronized even when PTS is simply predicted during video encoding.
[0096]
The number of pictures averaged by the PTS difference filter 25 is 6 pictures, which is a common multiple of 2 which is the picture period of the RFF flag and M = 3 which is the I and P picture periods. In this way, by setting the number of pictures averaged by the PTS difference filter 25 as a common multiple, the average value of the cumulative field number difference becomes constant in l / 6 fields (line 15), and stable synchronization processing can be performed. There is also an effect that can be done.
[0097]
Further, by setting the number of pictures averaged by the PTS difference filter 25 to the minimum value among the above-mentioned common multiples, that is, the least common multiple “6”, it is possible to perform synchronization processing with quick response. is there.
[0098]
【The invention's effect】
As described above in detail, according to the moving image compression encoding method and apparatus and the moving image compression encoding / decoding system of the present invention, a large amount of data for temporarily storing data output from the compression encoding means is used. Therefore, there is no need to provide a redundant memory, so that an inexpensive and efficient moving image compression encoding apparatus and moving image compression encoding / decoding system can be obtained.
[0099]
According to the moving image compression encoding method of the present invention as set forth in claim 1, the time information counted at the time of compressing and encoding the image data and the picture type information after the compression encoding are held, and the compression encoding process is performed. Since the time information corresponding to the same picture type information as the picture type information of the frame of the image data is selected and the time information is determined, in order to calculate the time information of the frame of a certain image data, Complicated processing such as calculating time information after detecting the RFF flag, which is frame field deletion information, becomes unnecessary.
[0100]
Furthermore, according to the moving image compression coding apparatus of the present invention as set forth in claims 2 and 3, the cumulative field number counted at the time of input to the compression coding means and the picture type information after compression coding are held. In addition, the time information can be calculated by a simple conversion of selecting the number of accumulated fields corresponding to the same picture type information as the picture type information output from the compression encoding processing means. This eliminates the need for complicated processing such as calculating time information after detecting the RFF flag, which is field deletion information for subsequent frames, in order to calculate time information for a frame of certain image data. It is possible to obtain a moving image compression encoding apparatus having a simple structure.
[0101]
Furthermore, according to the moving image compression coding apparatus of the present invention as set forth in claims 4 and 5, the time stamp counted at the time of input to the compression coding means and the picture type information after compression coding are held. In addition, the time stamp is determined by selecting a time stamp corresponding to the same picture type information as the picture type information output from the compression encoding processing means. This eliminates the need for complicated processing such as calculating the time stamp after detecting the RFF flag, which is the field deletion information for subsequent frames, in order to calculate the time stamp of a frame of certain image data. It is possible to obtain a moving image compression encoding apparatus having a simple structure.
[0102]
Furthermore, according to the moving picture compression encoding / decoding system of the present invention as set forth in claim 6, the RFF flag of the B picture appearing thereafter is detected in order to calculate the time stamp of the I and P pictures at the time of encoding. Then, it is possible to obtain a moving image compression encoding / decoding system that does not need to perform complicated processing such as calculating a time stamp.
[0103]
In addition, according to the moving image compression encoding / decoding system of the present invention as set forth in claims 7, 8 and 9, it is possible to obtain a moving image compression encoding / decoding system that performs a fast synchronization processing. .
[0104]
Further, according to the moving image compression encoding / decoding system of the present invention as set forth in claims 8 and 9, a moving image compression encoding / decoding system capable of performing stable synchronization processing at the time of encoding is obtained. Can do.
[Brief description of the drawings]
FIG. 1 is a block diagram illustrating a configuration example of a moving image compression coding apparatus according to the present invention in a first embodiment.
FIG. 2 is a flowchart showing a processing sequence of a field number counting circuit of the moving picture compression coding apparatus according to the first embodiment of the present invention.
FIG. 3 is a flowchart showing a processing sequence of a field number replacement circuit of the moving image compression coding apparatus according to the first embodiment of the present invention.
4 is a block diagram illustrating a configuration example of a moving image compression coding apparatus according to the present invention in Embodiment 2. FIG.
FIG. 5 is a block diagram illustrating a configuration example of a moving image compression encoding / decoding system according to the present invention in Embodiment 3.
FIG. 6 is a flowchart showing a processing sequence of a simple prediction circuit for the number of fields of a moving image compression encoding / decoding system according to the present invention in Embodiment 3;
FIG. 7 is a schematic diagram showing the cumulative number of fields predicted by the simple number-of-fields prediction circuit of the moving image compression encoding / decoding system according to the present invention in the third embodiment.
FIG. 8 is a schematic diagram showing the cumulative number of fields predicted by the simple number-of-fields prediction circuit of the moving image compression encoding / decoding system according to the present invention in the third embodiment.
FIG. 9 is a schematic diagram showing a procedure of 3: 2 pull-down processing.
FIG. 10 is a block diagram illustrating a configuration example of a moving picture compression encoding apparatus according to a conventional MPEG-2 system.
FIG. 11 is a schematic diagram showing a state before and after frame sequence encoding when a moving image compression encoding process according to the conventional MPEG-2 method is performed.
[Explanation of symbols]
5 video encoder, 6 field count circuit, 7 field number replacement circuit, 8 PTS conversion circuit, 9 PTS additional circuit, 10 90kHz frequency divider, 11 PTS count circuit, 12 PTS replacement circuit, 13 information extraction circuit, 14 field number Simple prediction circuit, 15 audio encoder, 16 PTS additional circuit, 17 multiplexing circuit, 18 recording device, 19 separation circuit, 20 audio decoder, 21 video decoder, 22 reference time generation circuit, 23 audio PTS comparison circuit, 24 video PTS comparison Circuit, 25 PTS differential filter, 26 SCR creation circuit, 50 inverse 3: 2 pull-down processor.

Claims (7)

By converting each frame of the original moving image into a moving image in which one frame is composed of two fields, and overlapping a specific field in each unit frame with a predetermined number of consecutive frames of the moving image as one unit. Detecting a duplicate field in an image signal obtained by adjusting a time axis in advance and deleting one redundant field;
Outputting field deletion information indicating that the field has been deleted;
Compressing and encoding the image signal after the redundant field is removed;
Wherein together with the image signal to count the cumulative number field plus Field Delete number to the number of times which is compression-encoded in frame units, defines the sequence order for another frame after the frame has been compression-encoded Outputting picture type information identifying the type of image data to be performed;
Storing the counted cumulative field number and picture type information;
Based on the output picture type information, the oldest cumulative field number among the cumulative field numbers corresponding to the compression-coded picture type information among the cumulative field numbers stored for each picture type information is sequentially selected. Steps,
Counting time information by converting the selected cumulative field number to the reference frequency clock number;
Deleting the selected cumulative field number;
Moving picture compression encoding method characterized by including the step of adding the counted time information to the frame of the compressed encoded image signal. By converting each frame of the original moving image into a moving image in which one frame is composed of two fields, and overlapping a specific field in each unit frame with a predetermined number of consecutive frames of the moving image as one unit. Detecting a duplicate field in an image signal obtained by adjusting a time axis in advance and deleting one redundant field;
Outputting field deletion information indicating that the field has been deleted;
Compressing and encoding the image signal after the redundant field is removed;
Generating a clock as a reference for a time stamp representing time information;
By counting the generated clocks, a time stamp at the time of compressing and encoding the image data in units of frames is calculated, and at the same time, the image data defining the arrangement order with respect to other frames after the frames are compressed and encoded Outputting picture type information identifying the type; and
An image after compression encoding among the time stamps storing the time stamp corresponding to the picture type information of the image data after storing the calculated time stamp and picture type information after compression encoding Sequentially selecting from the oldest time stamp among the same picture type information as the picture type information of the data;
Adding a selected time stamp to the data of each frame after compression encoding;
A moving image compression encoding method comprising: By converting each frame of the original moving image into a moving image in which one frame is composed of two fields, and overlapping a specific field in each unit frame with a predetermined number of consecutive frames of the moving image as one unit. A duplicate field in an image signal obtained by adjusting the time axis in advance is detected, one redundant field is deleted, field deletion information indicating that the field has been deleted, and an image after the redundant field is deleted A field thinning means for outputting a signal;
A compression encoding means for inputting an image signal and field deletion information output from the field thinning means, compressing and encoding the image signal, and outputting the number of compressed fields and field deletion information;
Outputs picture type information that specifies the type of image data that defines the arrangement order of each frame after the compression encoding of the image signal input to the compression encoding means with respect to other frames, and the field thinning means Field number calculating means for calculating the cumulative number of fields at the time of input to
The accumulated field number and the picture type information calculated by the field number calculating means are stored, and the accumulated field number corresponding to the picture type information output from the compression encoding means is stored. Field number selection means for sequentially selecting from the smallest cumulative field number in the same picture type information as the picture type information output by the compression encoding means;
Time stamp conversion means for converting the cumulative field number into a time stamp representing time;
A moving image compression coding apparatus comprising: time information addition means for adding a time stamp converted by the time stamp conversion means to data of each frame after compression encoding. By converting each frame of the original moving image into a moving image in which one frame is composed of two fields, and overlapping a specific field in each unit frame with a predetermined number of consecutive frames of the moving image as one unit. A duplicate field in an image signal obtained by adjusting the time axis in advance is detected, one redundant field is deleted, field deletion information indicating that the field has been deleted, and an image after the redundant field is deleted A field thinning means for outputting a signal;
A compression encoding means for inputting an image signal and field deletion information output from the field thinning means, compressing and encoding the image signal, and outputting the number of compressed fields and field deletion information;
Clock generating means for generating a clock that is a reference of a time stamp representing time information;
By counting the clocks generated by the clock generation means, a time stamp at the time when the image data is input to the compression encoding means in a frame unit is calculated, and another time after the frame is compression encoded A time stamp calculating means for outputting picture type information for specifying a type of image data defining an arrangement order with respect to a frame;
The time stamp calculated by the time stamp calculating means and picture type information are stored, and the compression code among the time stamps storing time stamps corresponding to the picture type information output from the compression encoding means is stored. A time stamp selecting means for sequentially selecting from the oldest time stamp among the time stamps of the same picture type information as the picture type information output from the converting means;
And a time stamp adding means for adding the time stamp selected by the time stamp selecting means to the data of each frame after the compression coding. By converting each frame of the original moving image into a moving image in which one frame is composed of two fields, and overlapping a specific field in each unit frame with a predetermined number of consecutive frames of the moving image as one unit. A duplicate field in an image signal obtained by adjusting the time axis in advance is detected, one redundant field is deleted, field deletion information indicating that the field has been deleted, and an image after the redundant field is deleted A field thinning means for outputting a signal;
An image signal output from the field decimation means and field deletion information are input, and an I-picture that is an intra-frame coded picture in which the image signal is independently compression-coded in units of frames, and the frame between the I picture Bidirectional predictive coding that is compression-encoded using a correlation characteristic between frames of a P picture that is a forward-predictive encoded picture that is compression-encoded using the correlation characteristics of the I and P pictures Compression encoding means for compressing and encoding a B picture that is a picture and outputting the number of compressed fields and field deletion information;
Image data that obtains a repetition period of an I-picture or P-picture after encoding of an image signal input to the compression encoding means and defines an arrangement order with respect to other frames after the frame is compression-encoded Information extracting means for extracting picture type information and field deletion information for specifying the type of
From the repetition period and picture type information of the I picture or P picture acquired and extracted by the information extraction means,
If the picture type is an I picture or P picture, the cumulative field number of the previous I picture or P picture is calculated from the number of fields corresponding to the repetition period and the B picture next to the previous I picture or P picture. Add the number of field deletions to the picture to make the cumulative field number of the picture,
When the picture type is a B picture and the previous picture type is an I picture or a P picture, “2” is added to the number of accumulated fields of the previous I picture or P picture and the previous I picture Or, by adding the B picture field deletion count from the B picture next to the P picture to the picture, the cumulative field number of the picture is obtained.
If the picture type is a B picture and the previous picture type is a B picture, add “2” and the number of field deletions of the picture to the cumulative number of fields of the previous B picture. A field number predicting means for the number of accumulated fields of a picture, a time stamp converting means for converting the accumulated number of fields into a time stamp of an image,
Image time stamp adding means for adding an image time stamp to the data after being compressed and encoded by the compression encoding means and outputting image packet data;
Voice time stamp adding means for adding voice time stamp to voice data and outputting voice packet data;
Multiplexing means for multiplexing and outputting the image packet data output from the image time stamp adding means, the audio packet data output from the audio time stamp adding means, and a reference time stamp;
Data separating means for separating the data multiplexed by the multiplexing means into image packet data and audio packet data;
Voice synchronization processing means for extracting voice time stamps from voice packet data and performing voice synchronization processing;
Voice decoding means for decoding and outputting voice packet data;
Time axis filter means for performing time axis filtering on the difference between the time stamp of the image extracted from the image packet data and the reference time stamp;
Image synchronization alignment processing means for performing image synchronization alignment processing with values output from the time axis filter means;
A video compression encoding / decoding system comprising: image decoding means for decoding and outputting image packet data.   The time axis filter means adds a difference between a time stamp of the image extracted from the image packet data and a reference time stamp for a predetermined number of pictures, and uses a quotient obtained by dividing the addition result by the predetermined number as a difference. 6. The moving image according to claim 5, wherein a filtering process is performed, and the predetermined number is a common multiple of a repeated picture period of field deletion information and a repeated picture period of an I picture or a P picture. Compression encoding / decoding system.   The time axis filter means adds a difference between a time stamp of the image extracted from the image packet data and a reference time stamp for a predetermined number of pictures, and uses a quotient obtained by dividing the addition result by the predetermined number as a difference. 6. The moving image according to claim 5, wherein filtering is performed, and the predetermined number is a least common multiple of a repeated picture period of field deletion information and a repeated picture period of an I picture or a P picture. Image compression encoding / decoding system.

Images