ES2368965T3 - INFORMATION TRAIN GENERATION DEVICE, INFORMATION TRAIN GENERATION PROCEDURE, CODING DEVICE, CODING PROCEDURE, RECORDING MEDIA AND PROGRAM FOR THEMSELVES. - Google Patents

Abstract

An information train generation apparatus that generates an information stream for intelligent reproduction, according to MPEG-4 AVC, which includes (i) OCGM memory management control order for managing a buffer memory (2004) that stores a decoded image as a reference image, and (ii) encoded images, said apparatus comprising: a generating unit (1004) operative to generate the encoded images by image coding, an operative order addition unit for adding the OCGM to a predetermined encoded image, characterized by: a rating unit (1002) operative to assess whether or not a first encoded image to which OCGM is added is a reference image B that can be referenced when another encoded image is decoded; and a repeating unit (1003) operative to generate a supplementary reinforcement information of decoded reference image markup ISR RMIRD when said valuation unit (1002) assesses that the first encoded image is the reference image B and added with OCGM; wherein said generating unit is operative to generate the information stream by adding the RMIRD ISR to a second encoded image I or a second encoded image P, so that said second image follows the first encoded image in decoding order.

Description

Information train generation apparatus, information train generation procedure, coding apparatus, coding procedure, recording medium and program for them

Technical field

The present invention relates to an information train generation apparatus that generates an information train that includes encoded images, an information train generation method, an image coding apparatus, an image coding procedure, a means of recording and a program for them.

Prior art

Document US-2002 / 0.016.970-A1 is related to a data conversion apparatus and procedure, a data distribution apparatus and procedure and a data distribution system, in particular in accordance with MPEG-2 video. The data distribution system includes a server and a decoding terminal. In the server apparatus, the data information of the access unit to be transmitted is read from a data storage unit in accordance with a request for intelligent reproduction. The information of the access unit is converted for intelligent reproduction in a data conversion unit. The data to be transmitted is converted into a multiplexing unit into intelligent reproduction data according to the information of the access unit that has been converted. Next, the intelligent playback data is transmitted from the server apparatus to the decoding terminal.

In the multimedia era that handles audio, video and other pixel values in an integrated manner, existing information media, specifically, newspapers, magazines, television, radio, telephone and the like through which information is transported to people, They have recently been included in the multimedia field. Generally, multimedia refers to something that is represented by the association not only of characters, but also of graphics, sound, and especially images and the like, together, but to include the information media mentioned above in the multimedia field, it becomes necessary as a prerequisite to represent such information in a digital form.

However, if the amount of information carried by each of the mentioned information media is estimated as the amount of digital information, while the amount of information for 1 character in the case of text is 1 to 2 bytes, the amount of information required for the sound is 64 Kbits per second (telephone quality), and 100 Mbits per second or more are necessary for motion pictures (current television reception quality), it is not realistic for the media to handle an amount of such huge information, since it is in digital form. For example, although videophones are already in real use through the Integrated Services Digital Network (ISDN) that offers a transmission rate of 64 Kbit / s to 1.5 Mbit / s, it is impossible to transmit images on television and images taken by cameras directly through ISDN.

Consequently, information compression techniques have become necessary, and for example, in the case of the videophone, standards H.261 and H.263 are being used for motion picture compression technology, internationally standardized by the International Union of Telecommunications — Telecommunication Standardization Sector (ITU-T). On the other hand, with the information compression techniques according to the MPEG-1 standard, it has also become possible to store video information on compact discs (CDs) of general music together with audio information.

In this case, Moving Picture Experts Group (MPEG) is an international standard for compression of motion picture signals standardized by the International Organization for Standardization and the International Electrotechnical Commission (ISO / IEC). MPEG-1 is a standard for compression of motion picture signals of up to 1.5 Mbps, in other words, for the compression of television signals up to about one hundredth. On the other hand, since the image quality object within the scope of the MPEG-1 standard is limited to a degree of quality of the medium that can be obtained through a transmission speed of approximately 1.5 Mbps, the use of MPEG-2 , which was standardized to meet the demands of improved image quality, makes the television broadcast quality with motion picture signals compressed up to 2 to 15 Mbps. In addition, MPEG-4, which has currently exceeded rates of compression of MPEG-1 and MPEG-2, and that also allows coding, decoding and operating on an object basis, and that performs the new functions required for the multimedia era, has been standardized by the working group (ISO / IEC JTC1 / SC29 / WG11) which has promoted the normalization of MPEG-1 and MPEG-2. MPEG-4 initially intended to normalize a low bit rate coding procedure. However, currently, it has been extended to the normalization of a more versatile coding procedure that also includes high bit rate coding for interlaced images. After that, MPEG-4 Advanced Video Coding (AVC) has been standardized as a new generation procedure for encoding images with a higher compression rate through a collaboration of ISO / IEC and ITU-T. It is intended to be used for devices related to new generation optical discs or for broadcast directed to mobile phone terminals.

Generally, in the encoding of a moving image, the amount of information is compressed by reducing redundancy in the temporal and spatial directions. Consequently, in a prediction coding between images intended to reduce temporal redundancy, an estimation of the movement and the generation of a predictive image is made on a block basis with reference to a previous or next image, and an encoding is encoded. difference value between the predictive image obtained and an image to be encoded. In this case, an image indicates a screening: it indicates a plot in a progressive image; and indicates a plot or field in an interlaced image. In this case, the interlaced image is an image whose plot consists of two fields that differ temporarily from each other. In an encoding and decoding of the interlaced image, it is possible to process a frame as a frame, process it as two fields or process it as a frame structure or as a field structure on a block basis in the frame.

An image I is an image that is intracoded without reference to a reference image. In turn, an image P is an image that is encoded with prediction between images with reference to a single image. In addition, an image B is an image that can be encoded with prediction between images with reference to two images at the same time. Image B may refer to two images as a pair of any images that are displayed before or after image B. For each block, which is a basic unit of encoding and decoding, a reference image may be specified. The reference image described above in an encoded bit stream is distinguished as a first reference image with the reference image described later as a second reference image. Note that, as a condition for encoding and decoding these images, it is necessary that the referenced image has already been encoded and decoded.

Fig. 1 is a drawing showing a structure of a conventional MPEG-2 information stream. As shown in Fig. 1, the MPEG-2 information stream has a hierarchical structure as described below. The information stream is made up of more than one Image Group (GDI), and random editing and access of a moving image is made possible by using the information stream as a basic unit for coding. Each GDI is made by more than one image. Each image is one between an image I, an image P or an image B. Each train of information, GDI and image are also made by synchronous code (sync) indicating a cut-off point of each unit and a heading that is the data common in the unit.

Fig. 2A and Fig. 2B are drawings showing an example of a predictive structure between images used in MPEG-2. In the drawings, the images shown shaded diagonally are images that other images will take as a reference. As shown in Fig. 2A, in MPEG-2, the image P (P0, P6, P9, P12, P15) can be encoded by prediction as a reference to an image I or image P that is displayed immediately before said image Q. In addition, image B (B1, B2, B4, B5, B7, B8, B10, B11, B13, B14, B16, B17, B19, B20) can be encoded by prediction as a reference to an image I or P image which is displayed before and after said image B. In addition, the order of configuration in an information stream has been determined as follows: images I and images P are configured in order of display; and each of the images B is configured immediately after an image I or image P that is displayed immediately after said image B. As a GDI structure, for example, as shown in Fig. 2B, the images from I3 to B14 They can be included in a GDI.

Fig. 3 is a drawing showing a structure of an MPEG-4 AVC information train. In MPEG-4 AVC, there is no concept equivalent to GDI. Therefore, in the case where the procedure for configuring parameter sets described below and the predictive structure of images are not limited, it is necessary to look for an image whose image data is analyzed sequentially and can be decoded when accessing They randomly. However, by separating the data into special image units according to which each image is decoded without relying on other images, it is possible to construct a unit that can be accessed randomly and that is equivalent to the GDI. These separate units are called random access units (UAA) and an information train that is constituted by UAA is called the information train that has a random access structure.

In this case, the access unit (hereinafter referred to as UA) that is a basic unit related to an information stream is explained. A UA is a unit used to store encoded data in an image, which includes parameter sets and sector data. The parameter sets are divided into a set of image parameters (CPI) which are data corresponding to a header of each image and a set of sequence parameters (CPS) corresponding to a heading of a GDI unit in MPEG-2 and more. The CPS includes a maximum number of images available for reference, image size and the like. The CPI includes a variable length coding procedure, an initial value of the quantization stage and a series of reference images. An identifier is linked to each image indicating that one is referenced between the CPI and the CPS.

For MPEG-4 AVC I images, there are two types of I images: a Decoder Instant Regeneration (RID) image; and an image I that is not the RID image. The RID image is an image I that can be decoded without reference to an image preceding the RID image in decoding order, i.e., whose condition necessary for decoding is reset, and is equivalent to a front image I of a closed MPEG GDI -2. For the image I that is not the RID image, an image that follows said image I in decoding order may refer to an image that is earlier than said image I in decoding order. In this case, the RID image and the image I indicate images constituted only by sectors I. The image P indicates an image constituted by sectors P or sectors I. Image B indicates an image constituted by sectors B, sectors P or sectors I. Note that the sectors of the RID image and the sectors of the non-RID image are stored in different types of CAR units.

The MPEG-4 AVC UA may include, in addition to data necessary for decoding an image, supplementary information called Supplemental Reinforcement Information (ISR) that is unnecessary for decoding an image, UA limit information and the like. Data such as the parameter set, sector data and the ISR are stored in a Network Abstraction Layer (CAR) unit (UCAR). The CAR unit consists of a header and a payload, and the header includes a field that indicates a type of data stored in the payload (hereinafter referred to as the CAR unit type). The value of the CAR unit type is defined for each type of data, such as a sector and an ISR. When referring to the type of CAR unit, you can specify the type of data stored in the CAR unit. The ISR CAR unit can store one or more ISR messages. The ISR message also consists of a header and a payload and a type of information stored in the payload is identified by a type of ISR message indicated in the header.

Fig. 4 is a drawing showing an example of a predictive structure of MPEG-4 AVC. In MPEG-4 AVC, an image P UA may refer to an image B UA. As shown in Fig. 4, the image P UA (P7) may refer to the image B UA (B2). In the present specification, in order to perform high-speed reproduction displaying only the UA of images I and images P, I0, B2, P4 and P7 must be decoded. Thus, when intelligent reproduction is performed, for example, in skip reproduction, variable speed reproduction or reverse reproduction, the UAs that must be decoded cannot be determined in advance so that in the end it is necessary to decode all UAs. However, if additional information indicating the UAs that need to be decoded for intelligent reproduction is stored in an information stream, the UAs to be decoded can be determined with reference to the supplementary information. Such supplementary information is called intelligent reproduction information. In addition, if a restriction in a predictive structure is set beforehand, for example, that the UA of P images does not refer to a UA of image B, only the UA of the images I and the P images can be decoded and displayed. In addition, for UA of images I and P images, UA of images I and P images can be decoded sequentially and displayed if the decoding order is the same as the display order.

Fig. 5 is a block diagram showing a structure of a conventional multiplexer.

A multiplexer 17 is a multiplexer that receives video data, encodes the video data received in MPEG-4 AVC information streams, generates database information about the encoded data, multiplexes and records the encoded data and the base information of data. It includes an attribute determination unit of the information stream 11, an encoding unit 12, a database information generating unit 13 having a general database information generating unit 14, a multiplexing unit 15 and a recording unit 16.

The attribute determination unit of the information stream 11 determines an encoding parameter for MPEG-4 AVC encoding and limited content in relation to intelligent reproduction, and sends it to the coding unit 12 as TYPE attribute information. In this case, the limited content related to intelligent reproduction includes information on whether or not to apply a limitation to construct a random access unit in an MPEG-4 AVC information stream, whether or not to include information that indicates to a UA that It will be decoded when variable speed or reverse reproduction is performed, or whether or not to impose a restriction on a predictive structure between UA. The encoding unit 12, based on the TYPE attribute information, encodes the input video data in an MPEG-4 AVC information stream, and sends the information stream access information to a general information generation unit of database 14 while sending the encoded data to the multiplexing unit 15. In this case, the access information indicates information about an access base that is a basic unit for accessing an information stream, which includes a starting address , size, time displayed and the like of a front UA at the access base. The attribute determination unit of the information stream 11 also sends information necessary to generate database information such as a compression procedure and a resolution such as general database information to the general information generation unit. database 14. The database information generation unit 13 generates database information, and consists exclusively of the general database information generation unit 14. The general database information generation unit of data 14 generates, with the access information and general database information, table data to which reference will be made when accessing an information stream and table data in which attribute information is stored, such as , for example, a compression procedure, and sends the generated table data to the multiplexing unit 15 as dat base information I INFO. The multiplexing unit 15 generates multiplexed data by multiplexing the encoded data and the INFO database information, and sends the multiplexed data to the recording unit 16. The recording unit 16 records the multiplexed data entered from the multiplexing unit 15 into an optical disc, a hard disk or a recording medium as a memory.

Fig. 6 is a block diagram showing a structure of a conventional demultiplexer.

A demultiplexer 27 is a demultiplexer that, in accordance with an externally entered order indicating that intelligent reproduction is performed, separates, decodes and displays the UA data of MPEG-4 AVC from the optical disk on which a train is recorded. MPEG-4 AVC information along with the database information. It includes a database information analysis unit 21, a decoding / display UA determination unit 23, a UA separation unit 24, a decoding unit 25 and a display unit 26.

The database information analysis unit 21 is constituted exclusively by the general database information analysis unit 22. An intelligent reproduction instruction signal is introduced to indicate that intelligent reproduction function is performed as speed reproduction variable, reverse playback or skip playback in the general database information analysis unit 22. When the intelligent reproduction instruction signal is input, the general database information analysis unit 22 analyzes the input signal obtaining ACS access information from the database information of the multiplexed data, obtains access destination information that includes address information of an access base in which a UA is to be decoded or displayed and the like, and notifies the separation unit of UA 24. The separation unit of UA 2 4 analyzes the UAs that constitute an access base, obtains the information of intelligent reproduction information about a UA that will be decoded and displayed, and sends the information obtained to the UA determination unit of decoding / visualization. The decoding / display UA determining unit determines a UA that will be decoded and displayed based on a predetermined rule, and notifies the identification information of the UA for decoding and the identification information of the UA that will be displayed respectively to the unit of separation of UA 24 and display unit 26. The separation unit of UA 24 separates the data in the UA for decoding based on the access destination information, and sends the separated data to the decoding unit 25. The unit decoding 25 decodes the input UA data, and sends the decoded data to the display unit 25. Finally, the display unit 26 selects a UA for which it is indicated that it will be displayed in the display UA information, and displays the selected UA. (See Japanese patent publication pending procedure 2003-18.549).

Description of the invention

In the decoding apparatus, reference images or images awaiting display are stored in a buffer called Decoded Image Buffer (TID) after decoding said images. However, the predictive structure of an image is flexible in MPEG-4 AVC so that memory management in the TID becomes complicated. Thus, there is the problem that it is difficult to perform intelligent reproduction as fast forward. For example, in the case where high speed reproduction is performed that only decodes and displays images I and P images, the memory management information can be stored in an image B that will be skipped. If the memory management information indicates saving a specific image in a TID without deleting it from the TID, the information cannot be obtained. Therefore, the specific image may be deleted from the TID and the following image I or image P referring to the specific image may not be able to be decoded.

Fig. 7A and Fig. 7B are diagrams showing TID memory management. As shown in Fig. 7A, image data for a plurality of frames can be stored in the TID. In this example, image data can be stored for four frames. Also, in the TID, two types of areas can be established: a long-term memory; and a short-term memory. Image data stored in short-term memories is poured sequentially from an image of the oldest decoding order. On the other hand, the image data stored in a long-term memory is stored in a TID and other images can refer to them until it is established that they cannot be referenced by other images, by means of a memory management order called Operation Memory Management Control (OCGM). For example, a long-term memory is used when it is necessary to store the image I for a longer time, such as when referring to a front image I in the random access base by the following images in decoding order. Note that images stored in short-term reports may also not be referenced by the OCGM. By default, each image is stored in a short-term memory. In this case, the memory management order can specify how many memories for how many frames are allocated respectively for long-term memory and short-term memory. Note that the memory management order can be directed exclusively to the reference images. Fig. 7B shows an example in which a memory is allocated for a frame for a long-term memory and memories are allocated for three frames for a short-term memory between a frame memory for four frames. The allocation of long-term memory and short-term memory can be dynamically changed depending on the order of OCGM memory management.

Fig. 8 shows an example of a use of the memory management order. Fig. 8 (a) shows an image configuration in a random access base. In the drawing, I, P, B and Br respectively indicate an image I, an image P, an image B without references and a reference image B. The number attached to each image indicates a display order. In this case, the image B without references indicates an image B to which no other image has referenced, and the reference image B indicates an image B to which other images have referenced. In addition, the arrows indicate predictive structures. For example, P9 indicates that it refers to P5 and I1, B2 refers to I1 and Br3, and Br3 refers to I1 and P5. P images only refer to images I or P images so that reference image B is not referenced. Fig. 8 (b) shows images shown in Fig. 8 (a), arranged in decoding order. In this case, in Br11, it is assumed that a memory management order to transfer I1 to a long-term memory is stored in header information of sectors that constitute an image. Fig. 8 (c) to (h) shows images stored in a TID when the TID can store image data for four frames. In this case, Br refers to only one image I or one image P immediately before or after Br in display order, and Br is deleted from the TID according to the memory management order of an image I or image P which are two images after Br in the order of display. Fig. 8 (c) shows images stored in the TID after deleting Br3 in P9. In the TID, I1, P5 and P9 are stored in short-term memory. After P13 is decoded, as shown in Fig. 8 (d), I1, P5, P9 and P13 are stored in the TID. In the present specification, TID is full since four images are stored. Then, after decoding Br11, Br11 must be stored in the TID. However, since the TID is full, it is necessary to delete one of the images stored in the TID. In this case, in principle of the TID I1 should be deleted, which was first decoded. However, it is shown that a long-term memory is allocated for I1 so that I1 is transferred to the long-term memory (Fig. 8 (e)). Consequently, when Br11 is stored, as shown in Fig. 8 (f), P5, which is decoded first, of the images stored in the short-term memories is deleted. Note that the deletion of P5 is also executed by the memory management order so that the memory management order stored in Br11 includes an order that specifically indicates that P5 is not referenced, in other words, indicates that P5 It can be deleted. Fig. 8 (g) shows a state in which Br11 is deleted and P17 is stored in the TID after decoding P17. Finally, while P21 refers to I1, I1 is transferred to a long-term memory when P21 is decoded and made available for reference so that I1 can be referenced without any problem (Fig. 8 (h)).

Next, a memory management problem is explained when intelligent reproduction is performed, such as fast forward and skip playback. Especially, fast forward reproduction by decoding and reproduction of only I images and P images (IP reproduction) has been introduced for general use. Fig. 9 shows the memory management when the random access unit that is the same as the one shown in Fig. 8, is reproduced in IP. First, after decoding of I1, P5 and P9, I1, P5 and P9 are stored in short-term TID memories as shown in Fig. 9 (c). Then, after storing P13, four of the I1, P5, P9 and P13 are stored in the TID so that the TID is filled (Fig. 9 (d)). Then, in principle, it is established that Br11 for a long-term memory to store I1 depending on a memory management order and I1 is transferred to a long-term memory. However, Br11 is skipped without being decoded so that I1 is still stored in short-term memory. Consequently, when the decoded P17 is stored in TID, I1 is deleted because it has the oldest decoding order among the images stored in the short-term memories (Fig. 9 (e)). Therefore, four images of P5, P9, P13 and P17 are stored in the TID when P21 is decoded so that P21 cannot be decoded because there is no I1 (Fig. 9 (f)). Thus, in the case where an image is decoded while skipping images when intelligent reproduction is performed, if an image is skipped in which a memory management order is stored, a memory management disorganization problem occurs. so that the following images cannot be decoded correctly. Therefore, IP reproduction cannot be performed without decoding of all reference images and the amount of processing according to IP reproduction has increased.

An object of the present invention is to provide an apparatus for generating information trains, a method for generating information trains and a recording medium that, when intelligent reproduction is performed, prevents the disorganization of intelligent reproduction because there is no the reference image necessary for decoding in a buffer memory, and easily performs intelligent reproduction of an image by jump playback.

In order to achieve the above object, an information train generation apparatus according to the present invention is provided as defined in claim 1.

According to this structure, the disorganization of intelligent reproduction can be prevented, which is caused because there is no reference image necessary for decoding in a buffer memory. In other words, intelligent reproduction can easily be done by jumping image playback.

In addition, the same units of a method of generating information trains and a recording medium are included in the present invention.

Additional information on the technical background of this application

The descriptions of the following Japanese patent applications including descriptive memory, drawings and claims are incorporated herein by reference in their entirety: Japanese Patent Application No. 2004-134.211 filed on April 28, 2004 and Application for Japanese Patent No. 2004

272,517 filed on September 17, 2004.

Brief description of the drawings

These and other objects, advantages and features of the invention will become apparent from the following description thereof taken in conjunction with the accompanying drawings illustrating a specific embodiment of the invention. In the drawings:

Fig. 1 is a drawing showing an information train structure in the MPEG-2 video.

Fig. 2A and Fig. 2B are drawings showing an example of a GDI structure in the MPEG-2 video.

Fig. 3 is a drawing showing an information train structure of MPEG-4 AVC.

Fig. 4 is a drawing showing an example of a predictive structure of MPEG-4 AVC.

Fig. 5 is a block diagram showing a structure of a conventional multiplexer that encodes an MPEG-4 AVC information stream and multiplexes the encoded information stream.

Fig. 6 is a block diagram showing a structure of a conventional demultiplexing apparatus that reproduces multiplexed data generated by the conventional multiplexer.

Fig. 7A and Fig. 7B are drawings indicating memory management in a decoded image / buffer in MPEG-4 AVC.

Fig. 8 is a drawing showing an example when a use of a memory management order is necessary.

Fig. 9 is a drawing showing memory management when images I and P are reproduced in a random access unit shown in Fig. 9.

Fig. 10 is a block diagram showing a structure of a coding apparatus according to a first embodiment.

Fig. 11 is a drawing showing a procedure for repeating a memory management order.

Fig. 12A is a drawing showing images and a memory management order when I-PA image is used in conventional technology.

Fig. 12B is a drawing showing images and a memory management order when using I-PA image

according to the first embodiment. Fig. 13 is an organization chart showing a coding procedure for performing memory management without cause disorganization when images I and P are reproduced.

Fig. 14 is an organization chart showing a coding procedure to perform memory management without

cause disorganization when decoding an I-PA image. Fig. 15 is a block diagram showing a decoding apparatus that performs a process of decoding according to a sixth embodiment.

Fig. 16 is an organization chart showing a decoding procedure of an encoded information stream that ensures that memory management can be performed without causing a disorganization when I and P images are reproduced.

Fig. 17 is a flowchart showing a decoding procedure of an encoded information stream that ensures that memory management can be performed without causing a disorganization when an I-PA image is reproduced.

Fig. 18 is a block diagram showing a structure of a first multiplexer according to a second embodiment.

Fig. 19A and Fig. 19B are diagrams showing content of supportive information of reproduction. Fig. 20 is a drawing showing a procedure for specifying a CAR unit in which the Reproductive support information.

Fig. 21 is an organization chart showing an operation of the first multiplexer.

Fig. 22 is a block diagram showing a structure of a second multiplexer according to a third embodiment. Fig. 23 is a block diagram showing a demultiplexer according to a fourth embodiment. Fig. 24 is an organization chart showing a first demultiplexer operation. Fig. 25 is an organization chart showing a second demultiplexer operation. Fig. 26 is a diagram showing hierarchy of HD-DVD data according to a fifth embodiment. Fig. 27 is a diagram showing a logical space structure in the HD-DVD. Fig. 28 is a diagram showing the structure of a VOB information file. Fig. 29 is an explanatory drawing of a time map. Fig. 30 is a diagram showing a playlist file. Fig. 31 is a diagram showing a structure of a program file that corresponds to the list of

reproduction.

Fig. 32 is a diagram showing a structure of a total disk database information file BC Fig. 33 is a diagram showing a structure of a file for recording a global event manager. Fig. 34 is a schematic block diagram showing the HD-DVD player according to the sixth form of

realization. Fig. 35A, Fig. 35B and Fig. 35C show recording media showing a program for

performing an image coding procedure and an image decoding procedure of the present invention.

Best embodiment of the invention

Hereinafter, embodiments of the present invention are explained with references to the drawings.

(First embodiment)

In this embodiment, an encoding apparatus and a decoding apparatus are explained that can obtain an order necessary to manage memories in a TID only from images necessary for skip reproduction when intelligent reproduction is performed.

The coding apparatus generates an information stream that includes a memory management order and encoded images. When the information stream is generated, the encoding apparatus assesses whether an encoded image to which a memory management order is added should be skipped or not when intelligent reproduction is performed, in the case where it is valued as an encoded image that must be skipped, repetition information indicating the same content as the order is added to an encoded image that should not be skipped and decoded after the encoded image to be skipped, when intelligent reproduction is performed.

Fig. 10 is a block diagram showing a structure of a coding apparatus 1000 in the present embodiment. The coding apparatus 1000 includes an image type determination unit 1001, a repeat evaluation unit 1002, a repetition information generating unit 1003, an image coding unit 1004 and an encoded data output unit 1005 The image type determination unit 1001 determines the image type of an image to be encoded and sends the determined image type Pt to the repeat evaluation unit 1002 and the image coding unit 1004. The coding unit of images 1004 encodes an input image Vin depending on the type of Pt image, sends the encoded pic data to the encoded data output unit 1005, and sends the OCGM memory management information to the repeat evaluation unit 1002. If not OCGM memory management information is sent for the encoded image, indicated in the OCGM memory management information. The repeat valuation unit 1002 assesses whether or not a memory management order should be repeated based on the OCGM memory management information and the image type Pt, and sends a valuation result to the information generation unit 1003 as a repeat order Re. The repeat information generation unit 1003 generates a RMIRD ISR when the repeat order Re orders to repeat the memory management order, and sends the ISR data to the data output unit coded 1005. In this case, when the Re repeat order orders to repeat the memory management order, the information necessary to generate the RMIRD ISR is also sent to the repetition information generating unit 1003. The data output unit encoded 1005 sends the encoded pic data and the ISR data.

Thus, the ISR data generated according to the repeat order Re includes the same content as the OCGM memory management information, and includes substantially a copy of the OCGM memory management information.

Fig. 11 shows a random access unit of an MPEG-4 AVC information stream stored in the information recording medium as an example of an information stream encoded by the coding apparatus 1000 in the present embodiment. Although the example is the same as the conventional example shown in Fig. 9, it differs in the conventional example in that a memory management order stored in Br11 is repeated in P17 using Enhanced Reference Image Markup Repeat Reinforcement Information Decoded (hereinafter referred to as ISR RMIRD). Specifically, in P17 a memory management order established in Br11 is repeated to transfer I1 to a long-term memory making P5 not a reference object. Consequently, even if Br11 is skipped when IP is played, it is known that I1 is transferred to a long-term memory in Br11 when P17 is decoded. Consequently, I1 is transferred to long-term memory, P5 is deleted from the TID after P17 is decoded, and P17 is stored instead (Fig. 11 (e)). Therefore, as shown in Fig. 11 (f), there is I1 in TID when P21 is decoded so that P21 can be decoded with reference to I1. Thus, when a memory management order is issued to a reference image B, the images I and the P images can be decoded without disorganizing the memory management even when IP reproduction is performed, repeating the memory management order using ISR RMIRD in an image P immediately after said reference image B in decoding order. In particular, the use of the reference image B is a significant feature of MPEG-4 AVC, in the random access base having a structure such as IB Br BPB Br BPB Br BP…, a quadruple speed reproduction can be easily performed by decoding of images I and P and double speed reproduction by decoding of images I, P and Br so as to increase the functional capacity of intelligent reproduction. In that case, it is very effective that memory management can be guaranteed without disorganization. In this case, when an image I is in a position other than the start of the random access base, the memory management order can be repeated in an image I immediately after said image I in decoding order using ISR RMIRD.

In this case, the memory management order addressed to Br may be repeated in an image that is different from an image P or image I immediately following said image in decoding order if it is guaranteed that there is an image to which said image makes reference in the TID when an image will be decoded in IP playback. For example, in the case where memory management is not disorganized even without being repeated in image P immediately after said image in decoding order, it can be transmitted to an image P subsequent to said image P. Also, it can be ensured that the Memory management is not disorganized when only the reference images I or P images are decoded.

In addition, the memory management order may be stored in an information stream encoded by information other than RMIRD or may be indicated separately as in the database information.

Furthermore, it can be ensured that memory management is not disorganized either when intelligent reproduction other than IP reproduction is performed. Hereinafter, an example is explained in which jump reproduction is performed. A jump playback is an operation to display images from an image at a specified time. When the image display starts from an image other than the front image in the random access unit, a necessary image is decoded visually for decoding an image that will be displayed from the front image of the random access unit . In this case, in MPEG-4 AVC, a reference relationship is flexible. Therefore, decoding processing for skip playback or reverse playback can be reduced using a P image in which a specific restriction for decoding is offered or by referring to an image (hereinafter referred to as Image-Access Point (PA) ). The I-PA image has the following two characteristics:

one.

The I-PA image can be decoded by selectively decoding I images or P images before the I-PA image in the order of decoding;

2.

An image located after the I-PA image in the decoding order does not refer to an image located before the I-PA image in the decoding order.

Fig. 12A is a drawing showing images and a memory management order when the I-PA image is used in conventional technology. In the drawing, an image shown as I-PA indicates an I-PA image. In order to decode I-PA25, it is only necessary to decode I1, P7 and P16 so that P4, P10, P13 and P22 can be skipped. Thus, by selective decoding of images, the number of images necessary to decode the I-PA image placed somewhere in the random access unit can be reduced. Consequently, decoding processing can be reduced when playback is performed midway in the random access unit. In addition, an image placed after I-PA25 in the decoding order does not refer to an image placed before I-PA25 in the decoding order. In addition, an image P necessary for decoding the I-PA image may be indicated in a stream of encoded information using an ISR message and the like or in database information. In this case, if an OCGM1 memory management order indicating that I1 is transferred to a long-term memory is stored in P4, the memory management order cannot be obtained when only one image necessary to decode I-PA25 is decoded. .

Fig. 12B is a drawing showing images and a memory management order when the I-PA image is used in the first embodiment. As shown in Fig. 12B, repeating an OCGM1 memory management order in P7 that is clearly decoded when I-PA 25 is decoded, it is found that it is necessary to store I1 in a long-term memory when P7 is decoded. Thus, in the case where a memory management order is issued to an image that will be skipped when the I-PA image is decoded, a memory management can be performed without disorganization by repeating the memory management order in a necessary P image. to decode the I-PA image. Note that, if it can be guaranteed that the memory management is not disorganized, the order can be repeated in the image P that is not immediately subsequent to the image P with the original memory management order although necessary to decode the I-PA image.

Furthermore, more generally, when viewing an image that is necessary to decode for decoding a specific P image, it can be ensured that memory management can be done by decoding only images that need to be decoded.

Fig. 13 is an organization chart of a coding procedure for generating a stream of encoded information in which a memory management is guaranteed that is not disorganized when IP reproduction is performed. Processing from step S1001 to step S1008 shows a processing to encode an image that constitutes a random access unit. First, in step S1001, it is assessed whether an image to be encoded is an image I or an image P. If it is image I or image P, the processing proceeds to step S1002, and otherwise Processing advances to step S1004. In step S1002, it is assessed whether a memory management order is issued to a reference image B that follows an image P or an image I immediately before the image to be encoded in decoding order. In the case where the memory management order is issued, the processing advances to step S1003, and advances to step S1004 if the order is not issued. In this case, where there is no reference image B immediately before the image to be encoded in decoding order in the random access unit as a front image in the random access unit, it is assessed that the order is not issued. . After this, in step S1003, an RMIRD ISR is generated in which the memory management order is stored. In the case where memory management orders are issued to a plurality of reference B images, the content of all memory management orders is included in the RMIRD ISR. Next, in step S1004, the image data is encoded and processing proceeds to step S1005. In step S1005, it is assessed whether or not a memory management order is issued to the current image. If the order is issued, the processing advances to step S1006, and advances to step S1008 if the order is not issued. In step S1006, it is assessed whether or not the current image is the reference image B. If the image is the reference image B, the processing advances to step S1007, and advances to step S1008 if it is not. In step S1007, the contents of the memory management order and information are stored to specify an image to which the memory management order is issued. Finally, in step S1008, the encoded data is sent. In this case, in which the RMIRD ISR is generated in step S1003, the output encoded data includes RMIRD ISR. Note that, in the case where an image type is not determined in step S1001, processing from step S1001 to step S1003 can be performed after step S1004. In addition, the encoded data of an image can be sent on an image-by-image basis, or can be sent sequentially when the encoding is complete.

Fig. 14 is a flow chart of a coding procedure for generating a train of encoded information in which memory management is guaranteed, which is not disorganized when I-PA is decoded. Although the fundamental processing is the same as in the processing for IP reproduction shown in Fig. 13, it differs in the titration processing in steps S1101, S1102 and S1003. In step S1101, it is assessed whether or not the current image is an I image or a P image necessary to decode the I-PA image. Next, in step S1102, it is assessed whether a memory management order is issued to an image P or an image I that follows the image P or image I necessary to decode an I-PA image immediately before the current image in order of decoding and it is unnecessary to decode the I-PA image, in the random access unit. Also, in step S1103, it is assessed whether or not the current image is image I or image P unnecessary to decode the I-PA image.

In this case, in which the I-PA image can be decoded by selectively decoding only the P image before the I-PA image, it can only be assessed if the P image is necessary to decode the I-PA image. However, in the case where it is necessary to decode an image I that is a header of a random access unit, it may be indicated that it is necessary to perform the decoding in image I.

In addition, the present procedure can be applied not only limited to I-PA images but also to images in general where there is a specific restriction in the predictive structure and the like.

Note that, by combining the processing shown in Fig. 13 and Fig. 14, memory management can be performed that is not disorganized when IP playback is performed and when I-PA decoding is performed. For example, an operation can be performed as an effective decoding of an image that will be skipped using the I-PA and starting IP playback from it.

In addition, in the case that an image that needs to be decoded when intelligent reproduction is performed is indicated in supplementary and similar information, a memory management order can be repeated so that a memory management order necessary for decoding is obtained by decoding only images necessary for decoding.

Fig. 15 is a block diagram showing a structure of a decoding apparatus 2000 in the present embodiment. The decoding apparatus 2000 includes an image type obtaining unit 2001, a decoding evaluation unit 2002, a management order analysis unit 2003, a TID 2004 and a decoding unit 2005. First, they are introduced the encoded data Vin in the image type obtaining unit 2001. The image type obtaining unit 2001 obtains an image type of an image by detecting an image limit between the encoded data Vin, and entering the image type Ptd in the decoding valuation unit 2002. The decoding valuation unit 2002 assesses, based on the type of image Ptd, whether or not to decode the image, and enters the result of the Rep valuation into the order analysis unit of management 2003 and the decoding unit 2005. The management order analysis unit 2003 executes the memory management processing if a memory management order is Repeats in the image data when instructed to decode the image based on the result of the Rep valuation, analyzing the repeated memory order (repetition information) and transmitting a Cmd management order to the TID. The decoding unit 2005, when instructed to decode the image based on the result of the Rep valuation, obtains the reference data Ref by issuing a Req request to obtain data from the reference image to the TID, decodes the PicDat image data obtained by the imaging unit, and sends the decoded image Vout. Note that an original memory order included in the sector data of an image must be executed by a unit that is not shown in the diagram.

Note that the supplementary information to specify an image that needs to be decoded when intelligent reproduction is performed can be stored in a train of information encoded as a front UA of the random access unit or in database information. In this case, the decoding valuation unit 2002 can determine a UA to decode by analyzing the supplementary information.

Fig. 16 is a flowchart showing a decoding operation of an encoded information stream in which a memory management is guaranteed that is not disorganized when IP reproduction is performed in the decoding apparatus 2000. First, in the step S2001, it is assessed whether an image to be decoded is an image I or an image P. When it is assessed that the image is image I or image P, the operation advances to step S2002. If the image is different from the above, the image processing ends without decoding the image and performs the processing of the next image. In step S2002, it is assessed whether or not the current image includes the RMIRD ISR, if it includes the RMIRD ISR, the operation advances to S2003, and otherwise, the operation advances to step S2004. In step S2003, the memory management processing is executed by analyzing the contents of the RMIRD ISR and the operation advances to step S2004. In step S2004, the image is decoded. Note that, in step S2003, the memory management processing is not performed if it has already been performed by the preceding order that is in the sector heading or in the RMIRD ISR.

Fig. 17 is an organization chart showing an operation when an I-PA image is decoded in a coded information stream in which a memory management is guaranteed that is not disorganized when the I-PA image is decoded. Although the fundamental processing is the same as the processing when IP reproduction is performed as shown in Fig. 16, it differs in the valuation processing in step S2101. In step S2101, it is assessed whether or not an image to be decoded is an image necessary to decode the I-PA image. If the image is necessary to decode the I-PA image, the operation advances to step S2002, and if it is not necessary, the image processing ends and the processing of the next image is performed.

When the memory management order is repeated using a procedure other than the RMIRD ISR, a memory management order is obtained by a predetermined procedure.

Note that, by combining the processing shown in Fig. 16 and Fig. 17, memory management can be performed that is not disorganized when IP reproduction is performed and when I-PA reproduction is performed.

In this case, in operations such as decoding of only the I images and the P images when IP reproduction is performed, or unnecessary P images or I images are skipped to decode the I-PA image when the I-PA image is decoded, the information of flag that guarantees that a memory management order necessary to manage the TID can be obtained from the image to be decoded can be adjusted to the database information or encoded information stream and the like. For example, in a Network Abstraction Layer (CAR) unit of a sector of the reference image B, a field value called nal_ref_idc that indicates whether or not the sector is a sector of the reference image is set to A value of one or more. In the non-reference image B, the same field is set to 0. Consequently, the nal_ref_idc field may be flag information. In addition, in the database information, a codec type information showing MPEG-4 AVC Video and MPEG-2 Video and the like can be used as a flag.

Note that, in the foregoing, the MPG-4 AVC is explained. However, a similar procedure can be applied to other coding procedures.

(Second embodiment)

Fig. 18 is a block diagram showing a multiplexer in the present embodiment.

A multiplexer 35 receives input video data, encodes the input data in an MPEG-4 AVC information stream, multiplexes and records the following information together with the information stream: UA access information that constitutes the information stream; and database information that includes supplementary information to determine an operation when intelligent reproduction is performed. The multiplexer 35 includes an information train attribute determination unit 11, an encoding unit 12, a database information generating unit 32, a multiplexing unit 34 and a recording unit 16. Equal marks are assigned to units that perform the same operations as in the conventional multiplexer shown in Fig. 5, and explanations on the same units are omitted in this case. Note that the coding procedure is not only limited to MPEG-4 AVC and other procedures such as MPEG-2 Video and MPEG-4 Video can be applied. In addition, it can include a coding unit 1000 instead of the coding unit 12.

The attribute determination unit of the information stream 11 determines an encoding parameter for MPEG-4 AVC encoding and a restriction related to intelligent reproduction, and sends it to the encoding unit 12 and a support information generating unit of reproduction 33 as attribute information TYPE. In this case, the restriction related to intelligent reproduction includes information on whether or not to apply a restriction to constitute a random access unit in an MPEG-4 AVC information stream, whether or not to include information indicating the UA which will be decoded or displayed when variable speed playback and reverse playback are performed, or whether or not to restrict the predictive structure between UA. The reproduction support information generation unit 33 generates, based on the TYPE attribute information entered, the HLP support information that indicates whether or not there is a random access structure, and sends the generated information to the unit of multiplexing 34. The multiplexing unit 34 generates multiplexed data by multiplexing the encoded data entered from the coding unit 12, database information INFO and supporting information HLP, and sends the multiplexed data to the recording unit 16. Note that the encoding unit 12 can send the MPEG-4 AVC information train in packets in an MPEG-2 Transport Train (TS), a Program Train (PS) and the like. Or you can package the information stream using a procedure defined by applications such as BD.

Fig. 19A and Fig. 19B show an example of information indicated in the HLP support information. The HLP support information has the following two procedures: a procedure for directly indicating information about an information stream as shown in Fig. 19A; and a procedure to indicate whether or not the information stream satisfies a restriction defined by a specific application standard as shown in Fig. 19B. In Fig. 19A, the following is indicated as information train information: i) if the information train has a random access structure; ii) if there is a restriction in a predictive structure between images stored in a UA; and iii) if there is information that indicates a UA to decode or a UA to display when intelligent reproduction is performed.

In this case, the information of the UA that is decoded or displayed when intelligent reproduction is performed can directly indicate the UA that is decoded or displayed, or it may indicate decoding or display priority. For example, the information that indicates the UA to decode or visualize in a unit of random access by unit can indicate that it is stored in a CAR unit that has a special type defined by the application. In this case, you can indicate if there is information that indicates a predictive structure between UA that constitutes a random access unit. In addition, the information about the UA that will be decoded or displayed when intelligent reproduction is performed can be added together for each group of more than one random access unit or can be added to each UA that constitutes a random access unit.

In addition, when the information indicating a UA to be decoded or displayed is stored in the CAR unit that has a special type, it can indicate a CAR unit type of the CAR unit. In an example of Fig. 20, in the HLP support information, the information about a UA that will be decoded or displayed when intelligent reproduction is performed is included in a CAR unit whose type of CAR unit is 0. In the present specification , by separating the CAR unit whose type of CAR unit is 0 from the UA data in the information stream, information relating to intelligent reproduction can be obtained.

In addition, the restriction of the predictive structure may indicate whether or not one or more predetermined restrictions are satisfied or if the following individual restriction is satisfied or not.

i) The respective UA of image I and image P have the same decoding and display orders.

ii) The UA of the image P does not refer to the UA of the image B.

iii) A UA displayed after a front UA in the random access unit only refers to UA included in the random access unit.

iv) Each UA can refer only to the maximum number N of UA in decoding order. In the present specification, one UA is counted for each UA reference or all UAs and the HLP support information may indicate the value of N.

Note that, in MPEG-4 AVC, in order to improve the image quality, as a reference image an image is used in which filtering (unlocking) is performed to reduce block distortion after decoding the image, while an image can be used before said unlocking as an image for display. In the present specification, it is necessary for the image decoding apparatus to store image data before and after unlocking. In this case, the information that indicates if necessary

or not storing the image before unlocking as an image for display can be stored in HLP support information.

In this case, the HLP support information may include all of the above information or may include some of the information. In addition, you can include the necessary information based on a predetermined condition. For example, it may include information on whether intelligent reproduction information exists only in the case where there is no restriction in a predictive structure.

In addition, the HLP support information may include information indicating the following: whether IP reproduction can be performed or not without causing disorganization of memory management by decoding only images I and images P; or if the I-PA image can be decoded or not without causing disorganization of memory management by decoding only the I images or P images necessary to decode the I-PA image.

In addition, information other than the above can be included in the HLP support information.

In Fig. 19B, the HLP support information does not directly show information relating to an information train structure but whether or not it satisfies the restrictions related to the information train structure defined by an HD-DVD standard which is a standard for storing high definition high definition (HD) image on a DVD. In addition, in an application standard such as BD-ROM, in the case where a plurality of modes are defined for restriction in the information stream structure, information indicating what mode is applied can be stored in the HLP supporting information. For example, it can be used so that a mode 1 has no restrictions, and a mode 2 has a random access structure and information is included in an information stream to specify a UA that is decoded when intelligent reproduction is performed. In this case, you can indicate whether or not the restrictions defined in a communication service such as continuous downloading and viewing, or a broadcast standard are satisfied.

Note that the HLP support information may indicate the information shown in Fig. 19A and in Fig. 19B. Also, in the case in which it is known that an information stream satisfies a restriction in a particular application standard, it may not indicate whether the information stream satisfies an application standard, but store the restriction on the application standard by becoming in a procedure to directly describe an information train structure as shown in Fig. 19A.

In this case, in which the information indicated in the HLP support information changes during the download in direct display, the information in each section can be stored respectively. For example, in the case where different information trains are edited and connected to each other, in the edited information train, the HLP support information may change during direct display download. Therefore, the content of the HLP support information is also switched.

Note that the information indicating a UA that will be decoded or displayed when intelligent reproduction is performed can be stored as database information.

Fig. 21 is a flow chart showing an operation of multiplexer 35. In a step s11, the TYPE attribute information is determined based on a user setting or a predetermined condition. In a step s12, an information stream is encoded based on the TYPE attribute information. In a step s13, the HLP support information is generated based on the TYPE attribute information. After that, in step s14, access information is generated for each access base of the encoded information stream, and INFO database information is generated along with other necessary information. In a step s15, the HLP support information and the INFO database information are multiplied, and the multiplexed data is recorded in a step s16. In this case, the operation of step s13 can be performed before the operation of step s12, or it can be performed after the operation of step s14.

(Third embodiment)

Fig. 22 is a block diagram showing a structure of a second multiplexer in the present embodiment.

A multiplexer 43 receives a packetized information stream that is distributed from a server that is not shown in the diagram, multiplexes and records, together with the information stream, general database information that includes access information to the UAs that make up a train of information and supplementary information to determine an operation when intelligent reproduction is performed. It includes a unit for obtaining attributes of the information stream 41, a unit for receiving the information stream 42, a database generating unit 32, a multiplexing unit 34 and a recording unit 16. The units are assigned same brands to units that perform the same operations as units in the multiplexer explained in the second embodiment, and explanations about them are omitted here.

The attribute obtaining unit of the information stream 41 generates TIPO attribute information based on information from the information stream obtained separately from the information stream, and sends the TYPE attribute information to the support information generating unit of reproduction 33. In this case, the information train information includes information relating to intelligent reproduction such as: if it should be applied

or not a restriction to constitute a random access unit in an MPEG-4 AVC information stream; whether or not to include information indicating the UAs that will be decoded or displayed when performing variable speed playback and reverse playback; and whether or not a restriction should be given in a predictive structure between UA. The information train reception unit receives an MPEG-4 AVC information train, which is packetized by an MPEG-2 Transport Train (TS) and a Real Time Transmission Protocol (RTP), sends the received information train to multiplexer 34 as an information stream for recording, and also sends the access information to the general database information generating unit 14.

In this case, when the TS packet, the RTP packet and the like are received in an environment where packet loss is generated, in the case where error concealment processing is performed when the information and data indicate that they have been lost data in an information stream due to packet loss, the HLP can store information about this as supporting information. As information indicating the loss of data, the following information may be displayed: flag information indicating whether the data in an information stream has been lost or not; information indicating that a special error notification code is entered in an information stream in order to notify the lost part; or identification information of an error notification code to be inserted.

(Fourth embodiment)

Fig. 22 is a block diagram showing a structure of a demultiplexer in the present embodiment.

A demultiplexer 55 separates an MPEG-4 AVC information stream from the multiplexed data generated by the multiplexers explained in the second and third embodiments, and reproduces the separate MPEG-4 AVC. It includes a database information analysis unit 51, an intelligent reproduction operation determination unit 53, a decoding / display UA determination unit 54, a UA 24 separation unit, a decoding unit 25 and a display unit 26. In this case, the same marks are assigned to the units that perform the same operations as the units in the conventional demultiplexer shown in Fig. 6 and explanations for the same units are omitted.

The database information analysis unit 51 includes supporting information of the reproduction analysis unit 52 and a general database information analysis unit 22. The supporting information of the reproduction analysis unit 52 When an intelligent reproduction instruction signal is introduced, it obtains and analyzes the HLP support information from the database information in the multiplexed data, generates intelligent reproduction support information based on the analysis result, and notifies to the intelligent reproduction operation determination unit 53 the supporting information of the intelligent reproduction. The intelligent reproduction operation determination unit 53 determines a procedure for determining a UA that will be decoded and displayed when intelligent reproduction is performed based on the supporting information of the intelligent reproduction, and notifies the decoding UA determination unit / display 54 an intelligent playback mode indicating the determined procedure. The decoding / display UA determination unit 54 analyzes the intelligent reproduction information TRK obtained by the separation unit of UA 24, determines a UA that will be decoded and displayed by a procedure indicated by the intelligent playback mode MODE, and notifies to the separation unit of UA 24 and the display unit respectively of the identification information of the UA for decoding and of the identification information of the UA to be displayed. In this case, the UA to be displayed can be determined by the decoding / display UA determining unit 54 based on a specified reproduction rate and the like. In addition, when the TRK intelligent reproduction information is stored in the database information, the TRK intelligent reproduction information stored in the database information can be obtained by setting another unit in the database information analysis unit 51 .

Fig. 24 is a flow chart showing operations of demultiplexer 55. When an intelligent reproduction instruction signal is input, it obtains HLP supporting information from the multiplexed data in step s20. In step s21, an operation for determining a UA that will be decoded and displayed based on the HLP support information obtained. In a step s22, it is assessed whether or not to use the TRK intelligent reproduction information when intelligent reproduction is performed. In a step s23, the intelligent reproduction information TRK is obtained from the information stream and analyzed, and the operation advances to a step s24. If the TRK smart playback information is not used, the operation proceeds directly to step s24. In step s24, a UA is determined to be decoded and displayed based on the procedure determined in step s21, and the operation proceeds to a step s25. In step s25, the determined UA is decoded and displayed. Note that the HLP support information can be obtained only in the case where intelligent reproduction is performed when playback starts or after playback starts.

Fig. 25 is an organization chart showing the content of the processing in step s21. Thereafter, the evaluation in steps s30, s33, and s35 is performed based on the intelligent reproduction support information obtained from the HLP support information. In step s30, it is assessed whether or not the information train has a random access structure, the operation advances to step s31 if the information train has a random access structure, and the operation advances to step s32 if the Information train does not have the random access structure. Next, the respective UA is determined to be decoded. In step s31, it is determined to start decoding from a front UA in the random access unit. In step s32, it is determined to start decoding from said UA when the front UA in the random access unit is a UA of a RID image, the decoding is initiated from said UA. In this case, in which a display time of a UA that includes the preceding RID image is before a predetermined time or more, a UA for decoding first can be determined based on a predetermined rule such as starting the decoding of a front UA in No previous access base or an image I other than RID. In step s33, it is assessed whether or not TRK intelligent reproduction information is included in the information stream. If the TRK is included, the operation advances to a s34 stage, and if the TRK is not included, the operation advances to a s35 stage. In step s34, the processing ends by determining a UA that will be decoded or displayed based on the TRK intelligent reproduction information. In step s35, it is assessed whether or not there is a restriction in the predictive structure between UA. If there is a restriction, the operation advances to a s36 stage, and if there is no restriction, the operation advances to a s37 stage. In step s36, the processing ends by determining that only one UA should be decoded when the UA is decoded which is necessary to visualize when variable speed reproduction and reverse reproduction is performed based on the restriction in the predictive structure. Also, in step s37, the processing ends by determining that all UAs are decoded. Consequently, a procedure for determining a UA to decode as a consequence of steps s31 and s32 is first determined, and a procedure for specifying a UA to decode when variable speed reproduction or intelligent reproduction is performed as a result of steps s34 , s36 and s37. Next, the decoding / display UA determination unit 54 is sent as information on the respective modes of intelligent playback modes MODE. Note that, when skip playback is performed, the processing can be terminated after step 32 or step s31. In this case, as a procedure for determining a UA to decode when intelligent reproduction is performed, a predetermined procedure can be used: for example, in the case where it is assessed that the TRK intelligent reproduction information is not included in the train of information in step s33, only UAs of images I and P images are decoded, or only UAs of images I, images P and images B referred to are decoded.

Note that, in the case where information is included to determine a UA that will be displayed in the TRK smart play information, information indicating the determination of a UA to be displayed based on the TRK smart play information can be included in the Smart play mode MODE.

In this case, where decoding cannot be performed by a procedure determined by the intelligent reproduction operation determination unit 53, a UA can be determined to decode by a predetermined procedure. For example, in the case where it is indicated to obtain the TRK reproduction information from the encoded information stream by the intelligent reproduction mode MODE, if the TRK intelligent reproduction information cannot be obtained in the encoded information stream, it can be determined that all UAs or UAs that will be decoded are decoded based on other information obtained from the HLP support information. In the present specification, it can be verified whether the database information includes the intelligent reproduction information TRK.

In this case, where information other than information related to intelligent reproduction is included in the HLP support information, a decoding or display operation can be switched according to the information. For example, operations can be switched based on packet loss information when data received is recorded through broadcast and communication.

In addition, a medium in which multiplexed data is recorded is not only limited to an optical disc but can also be another recording medium such as a hard disk and non-volatile memory.

In addition, the operations of the decoding / display UA determination unit 23 differ from each other. By preparing the conventional demultiplexers shown in Fig. 6, a demultiplexer to be used can be switched based on an intelligent reproduction mode determined by supporting information set separately from the reproduction analysis unit 52 and the reproduction operation determination unit. intelligent 53. For example, either of the two conventional demultiplexers having the decoding / display UA determination unit 23 performing the following three types of operations is prepared, and the demultiplexer to be used based on the supporting information can be switched HLP of the multiplexed data to be reproduced. The three types are: i) determine a demultiplexer so that all UAs are decoded throughout the time; ii) obtain the information of intelligent reproduction TRK during all the time and determine a UA to decode; and iii) determine a UA to decode assuming that an information stream follows a specific predictive structure.

(Fifth embodiment)

As a method of recording multiplexed data on an optical disk by a multiplexer according to the second embodiment, a method of storing HLP supporting information is explained as database information of a BD that is a new generation optical disk.

First, a recording format of a BD-ROM is explained.

Fig. 26 is a diagram showing a structure of a BD-ROM, in particular a structure of a BD disk

(104) which is a disk medium and data (101, 102, and 103) recorded on the disc. The data that will be recorded on the BD disk (104) is AV data (103), database information BD (102) as database information relative to AV data and UA playback sequence, and a program of BD playback (101) for performing an interactive. In the present embodiment, a BD disc is mainly explained for an AV application for playing AV content in a movie for the purpose of explanation. However, there is no doubt that it is the same even if it is used for other uses.

Fig. 27 is a diagram showing a structure of logical data directories / files recorded on the BD disc. The BD disc has a spiral recording area from the inner radius to the outer radius, for example, similar to DVD, CD and the like, and has a logical address space in which logical data can be recorded between a part front of the inner radius and a rear of the outer radius. Also, there is a special area inside the front that is read only by a controller called Burst Cut Area (BCA). This area cannot be read from the application so that it can be used, for example, for copyright protection technology and the like.

In the logical address space, application data is recorded as front of video data using the final system information (volume). As explained in conventional technology, the file system is UDF, ISO96660 and the like. It allows the reading of logical data stored as in the PC personal computer, using a directory and file structure.

In the present embodiment, such as the directory structure and files of the BD disk, a BDVIDEO directory is placed immediately below the root directory (ROOT). In this directory, the data (101, 102 and 103 explained in Fig. 26) are stored as AV content and database information processed in the database.

Under the BDVIDEO directory, the following seven types of files are recorded:

BD.INFO (fixed file name)

A file is recorded in the file that is one between "BD database information" and information relative to the BD disk as a whole in the file. A BD player reads this file first.

BD.PROG (fixed file name)

A file that is one between "BD playback program" and playback control information relative to the BD disc as a whole is recorded in the file.

XXX.PL (“XXX” is variable, the extension “PL” is fixed)

A file is recorded in the file that is one between “BD database information” and information list of reproduction that is a scenario (sequence of reproduction). There is a file for each playlist. XXX.PROG (“XXX” is variable, the extension “PROG” is fixed) A file that is one between “BD playback program” and control information is recorded in the file.

Playback for each playlist. A correspondence with a playlist is identified by a file body name (“XXX” correspondence).

YYY.VOB ("YYY" is variable, the extension "VOB" is fixed) A file is recorded in the file that is one between “AV data” and VOB (same as VOB explained in the example conventional). There is a file for each VOB.

YYY.VOBI (“YYY” is variable, the extension “VOBI” is fixed) A file is recorded in the file that is one between “BD database information” and train information of

VOB database information that is AV data. A correspondence with a VOB is identified by a file body name (correspondence "YYY"). ZZZ.PNG (“ZZZ” is variable, the extension “PNG” is fixed) A file that is one between “AV data” and PNG image data (an image format normalized by W3C, and

pronounced "ping") for structuring subtitles and a menu in the file. There is a file for each image

PNG With references to Figs. 28 to 32, a BD navigation data structure (base information of BD data).

Fig. 28 is a diagram showing an internal structure of a VOB database information file

("YYY.VOBI"). The VOB database information has attribute information of the information stream (Attribute) of said VOB and a time map (TMAP). There is an information stream attribute for each video attribute (Video) and audio attribute (Audio # 01 to Audio # m). In particular, in the case of audio train audio information, such as the VOB can have audio information trains at the same time, the number of audio information trains (Number) indicates whether a data field exists or not.

The video field attribute (Video) and its values are indicated below.

Compression procedure (Coding): MPEG1 MPEG2 MPEG3 MPEG4 (Advanced Video Coding)

Resolution: 1920 x 1080 1440 x 1080 1280 x 720 720 x 480

720 x 565 Width-height ratio

4: 3

16: 9

Frame rate 60

59.94 (60 / 1,001) fifty 30

29.97 (30/01001) 25 24

23,976 (24 / 1,001)

The following are the fields of the audio attribute (Audio) and their values.

Compression procedure (Coding):

AC3

MPEG1

MPEG2

LPCM

Number of channels (Ch):

1 to 8

Language Attribute (Language)

A time map (TMAP) is a table that has information for each VOBU. The table includes the VOBU number (Number) saved by that VOB and each VOBU information (VOBU # 1 to VOBU # n). Each of the VOBU information consists of an I # address beginning of an address of a forward VOBU TS packet (image start I), a travel direction (I # end) to the end address of the image I, and a playback start time (PTS) of the image I. In the case where an MPEG-4 AVC information train has a random access structure, a VOBU corresponds to one or more random access units.

Fig. 29 is a diagram explaining the details of the VOBU information.

As is widely known, there is a case in which an MPEG video information stream is compressed at variable bit rate in order to record in high image quality, there is no simple correspondence between the playback time and the size of the images. data. On the contrary, an AC3, which is an audio compression standard, compresses the audio at a fixed bit rate. Therefore, a relationship between a time and a direction can be obtained through a primary expression. However, in the case of MPEG video data, each frame has a fixed display time, for example, in the case of NTSC, a frame has a display time of 1 / 29.97 seconds. However, the data size of each compressed frame changes greatly depending on the characteristic of the image and the type of image used for compression, in particular I / P / B images. Therefore, in the case of MPEG video, a relationship between time and direction cannot be described in a primary expression.

In fact, it is impossible to describe an MPEG system information stream that is obtained by multiplexing MPEG video data in the form of a primary expression. Specifically, VOB also cannot describe time and size of data in a primary expression. Therefore, a time map (TMAP) is used to establish a relationship between time and direction in the VOB.

Therefore, when time information is given, first, the VOBU to which the time belongs belongs (PTS track for each VOBU), the PTS is skipped immediately before that time to a VOBU that has a TMAP (address specified by I # start), the decoding of images from a front image I in the VOBU is started, and the display of images from an image of that time starts.

Next, with reference to Fig. 30, the internal structure of a play information list ("XXX.PL") is explained.

The play information list consists of a list of cells (Cell List) and a list of events (Event List).

The cell list (Cell List) is a play sequence in a playlist, and the cell is played in an order of description in the list. The content of the cell list (Cell List) includes the number of cells (Number) and the information of each cell (Cell # 1 to Cell # n).

The cell information (Cell #) includes a VOB file name (VOB Name), a start time (Input) and an end time (Output) in the VOB, and a subtitle table. The start time (Input) and the end time (Output) are described by frame numbers in each VOB, and an address of the VOB data required for reproduction can be obtained using the time map (TMAP).

The subtitle table is a table that has subtitle information that will be played at the same time with the VOB. The subtitle can have a plurality of languages similar to audio, and the first information in the subtitle table is made up of a number of languages (Number) and the following table for each language (Language # 1 to Language # k).

Each language table (Language #) consists of language information (Language), the subtitle information number (Number) that is displayed separately, and subtitle information of the subtitle (Voice # 1 to Voice # j). The subtitle information (Voice #) consists of a corresponding image data file name (Name), a subtitle display start time (Input), a subtitle display end time (Output) and a subtitle display position (Position).

The event list (Events List) is a table in which the events generated in the playlist are defined. The list of events consists of each event (Event # 1 to Event # m) after the number of events (Number). Each event (Event #) consists of an event type (Type), an event ID (ID), an event generation time (Time) and a duration.

Fig. 31 is an event manager table (“XXX.PROG”) that has an event manager (time event and user event for menu selection) for each playlist.

The event manager table has an event manager / number of programs (Number) defined and an individual event / program manager (Program # 1 to Program # n). The description in each event manager / program (Program #) has a definition on the start of the event manager (<event_handler> tag) and an ID of an event manager linked to the event. After that, the program is described between keys “{” and “}” after a Function. Events (Event # 1 to Event # m) stored in an event list of “XXX.PL” are specified using an ID of an event manager.

Next, with reference to Fig. 32, the internal information structure related to a BD disk as a whole ("BD.INFO") is explained.

The total information of the BD disc consists of a list of titles and an event table for a global event.

The list of titles consists of a number of titles on a disc (Number) and then the information of each title (Title # 1 to Title # n). The information for each title (Title #) includes a playlist table (TablePL) included in a title and a list of chapters in the title. The playlist table (TablaPL) includes the number of playlists in the title (Number) and a playlist name (Name), specifically, a file name of the playlist.

The chapter list consists of a number of chapters (Number) included in the title and information of each individual chapter (Chapter # 1 to Chapter # n). Each fragment of chapter information (Chapter #) has a table of cells included in the chapter. The cell table consists of a number of cells (Number) and input information for each individual cell (Cell Input # 1 to Cell Input # k). The cell input information (Cell Input #) is described by a playlist name that includes the cell and a cell number in the playlist.

The event list (Event List) has a number of global events (Number) and information about each individual global event. In this case, it should be mentioned that a global event defined first is called the first event, which is an event that will be invoked first when a BD disc is inserted into a player. The event information for the global event has only one type of event (Type) and event ID (ID).

Fig. 34 shows a table of a program of a global event manager ("BD.PROG"). This table is the same as the event manager table explained in Fig. 32.

In said BD-ROM format, the HLP support information is stored as attribute information of the information stream of the VOB database information. When HLP support information is used only for MPEG4 AVC, HLP support information can be stored only when the compression procedure is MPEG-4 AVC.

Note that, in addition to the attribute information of the information stream and a time map, the HLP support information can be stored by establishing an area for storing the reproduction support information in the VOB database information. In addition, the HLP support information can be stored as BD database information other than VOB database information.

In addition, the HLP support information can be stored not only in the BD-ROM format but also in other recording formats such as BD-RE (Rewritable) as database information.

(Sixth embodiment)

Fig. 34 is a block diagram showing approximately the functional structure of a player that reproduces data recorded on the BD disc according to the fifth embodiment.

Data recorded on a BD disc (201) is read through the optical sensor (202). The read data is transferred to a special memory depending on the respective data type. The BD playback program (contents of the files “BD.PROG” or “XXX.PROG”), the database information BD (“BD.INFO”, “XXX.PL” or “YYY.VOBI”), and the AV data ("YYY.VOB" or "ZZZ.PNG") is transferred respectively to a program recording memory (203), a database information recording memory (204) and an AV recording memory (205).

The BD playback program recorded in the program recording memory (203), the BD database information recorded in the database information recording memory (204) and the AV data recorded in the AV recording memory (205) are processed respectively by a program processing unit (206), a database information processing unit (207) and a presentation processing unit (208).

The program processing unit (206) processes a program to receive information about the playlists that will be reproduced by the database information processing unit (207) and event information such as the execution time of a program . In addition, the program can dynamically change the playlists that will be played. In this case, it can be done by sending a play instruction from the playlist to the database information processing unit (207). The program processing unit (206) receives an event from a user, specifically, a request sent from a remote controller key, and executes the event if there is a program corresponding to the user event.

The database information processing unit (207) receives an instruction from the program processing unit (206), analyzes the corresponding playlist and the VOB database information corresponding to the playlist, and orders that the object AV data be reproduced to the presentation processing unit (208). In addition, the database information processing unit (207) receives standard time information from the presentation processing unit (208), orders the presentation processing unit (208) to stop the playback of AV data based on the time information, and also generates an event that indicates a program execution time for the program processing unit (206).

The presentation processing unit (208) has decoders that correspond respectively to video, audio and subtitle / image (still image). Each of the decoders decodes AV data based on an instruction sent from the database information processing unit (207), and sends the decoded AV data. The video, subtitle and image data are described respectively in a special plane, a video plane (210) and an image plane (209) after decoding, and synthesize the images by means of the synthesizing unit (211) and the They send to a display device like a television.

The following explains player operations when smart playback is performed.

The database information processing unit 207 includes a function of the intelligent reproduction operation determination unit 53 in the demultiplexer 55 according to the fourth embodiment, when an intelligent reproduction instruction signal is introduced to perform intelligent reproduction as variable speed reproduction, reverse reproduction or skip reproduction by means of the program processing unit 206, obtains and analyzes the HLP support information from the database information memory 204, and determines a procedure for determining the decoding and display operation when intelligent reproduction is performed. The presentation processing unit 208 includes a function of the decoding / display UA determining unit 54 in the demultiplexer 55, determines a UA to be decoded and displayed based on the procedure determined by the base information processing unit of 207 data, and decodes and displays the determined UA. In this case, the database information processing unit 207 may have the function of the decoding / display UA determination unit 54.

In addition, when the TRK intelligent reproduction information is stored in the database information BD, the database information processing unit 207 obtains the intelligent reproduction information TRK from the memory database information 204. The TRK intelligent reproduction information obtained is analyzed in the presentation processing unit 208.

Note that each function block in the block diagram shown in Figs. 10, 15, 18, 22 and 23 can be carried out as an LSI that is an integrated circuit apparatus. Said LSI can be incorporated into one or several chip forms (for example, function blocks other than a memory can be incorporated into a single chip). In this case, LSI is taken as an example, although it can be called “IC”, “LSI system”, “super LSI” and “ultra LSI” depending on the degree of integration.

The procedure for incorporation into an integrated circuit is not limited to LSI, and can be done with a general or private line processor. After LSI manufacturing, Field Programmable Gate Array (FPGA) can be used which is a programmable, or reconfigurable processor that can reconfigure the connection and settings for the circuit cell in LSI.

In addition, together with the arrival of the technique for incorporation into an integrated circuit, which replaces the LSI due to progress in semiconductor technology or another technique that has been derived from it, the integration of the function blocks can be carried out using the New technologies. The application of biotechnology can be cited as one of the examples.

Among the function blocks, only one unit can be constructed to store data separately without incorporating them in chip form, such as the storage medium 115 described in the present embodiment.

Note that the main part in the function blocks shown in Figs. 10, 15, 18, 22 to 25 and 34 or in the flowcharts shown in Figs. 13, 14, 16 and 17 can be carried out by means of a processor or a program .

As stated above, it is possible to employ the image coding procedure and the image decoding procedure presented in the above embodiment in any of the devices and systems described above. Consequently, it becomes possible to achieve the effects described in the aforementioned embodiment.

(Seventh embodiment)

In addition, when registering a program for the provision of the motion image coding procedure or the motion image decoding procedure as shown in each of the above-mentioned embodiments, on a recording medium such as a floppy disk , it becomes possible to easily perform the processing as shown in each of the above embodiments in an independent computer system.

Fig. 35A, Fig. 35B and Fig. 35C are diagrams of a recording medium for recording a program for performing the motion image coding procedure and the motion image decoding procedure in the above embodiments. in the computer system.

Fig. 35B shows the front view of a flexible disk and the schematic cross-section, as well as a flexible disk itself, while Fig. 35A shows an example of a physical format of the flexible disk as a recording medium itself. A flexible disk FD is contained in a box F, a plurality of tracks Tr are formed concentrically on the surface of the disk in the radial direction from the periphery, and each track is separated into 16 sectors Se in the angular direction. Therefore, on the floppy disk that stores the aforementioned program, the previous program is recorded in an area assigned for it on the previous FD floppy disk.

In addition, Fig. 35C shows the settings for recording and playing the program on and from the FD floppy disk. When the program is recorded on the FD floppy disk, the Cs computer system writes in the motion image encoding procedure and the motion image decoding procedure as the program on the FD floppy disk by means of a floppy disk drive . When the motion picture coding procedure and the previous motion picture decoding procedure are constructed in the computer system using the program recorded on the floppy disk, the program is read from the floppy disk by means of the floppy disk drive and is transferred to the computer system.

Note that the above explanation is offered assuming that a recording medium is a floppy disk, but the same processing can also be performed using an optical disk. In addition, the recording medium is not limited to these, but any other medium such as a CD-ROM, a memory card and a ROM cassette can be used in the same way if a program can be recorded on them.

Although only some illustrative embodiments of this invention have been described in detail above, those skilled in the art will readily appreciate that many modifications are possible in the illustrative embodiments without materially departing from the new teachings and advantages of this invention. Accordingly, it is intended that all of these modifications be included within the scope of this invention.

Industrial applicability

A multiplexer and a demultiplexer according to the present invention can perform decoding or display. effective when the data obtained by multiplexing a train of MPEG-4 information is reproduced in a special way AVC, so that the present invention is especially effective for a media playback device Package that focuses on a smart playback function.

Below is a list of additional embodiments of the invention:

Embodiment 1. An information train generation apparatus that generates an information train that includes encoded images and an order to manage a buffer that saves a decoded image as a reference image, the order being added to one of the encoded images, said apparatus: an operational assessment unit to assess whether the encoded image to which the order is added should be skipped or not at the time of intelligent reproduction; an operative addition unit to add, in the case where it is judged that the encoded image must be skipped, Repeat information indicating the same content as the order to another encoded image that follows, in order decoding, to the encoded image that is valued that should be skipped and that is not skipped at the time of smart reproduction; Y an operational generation unit to generate the information stream that includes the encoded images, the Order and repeat information.

Embodiment 2. The apparatus for generating information trains with the characteristics of the form of embodiment 1, in which the order orders to change an attribute of the reference image stored in the buffer from a short term memory to a long term memory.

Embodiment 3. The apparatus for generating information trains with the characteristics of the form of

embodiment 1, wherein said valuation unit is operative to assess that a reference image B is skipped in the moment of intelligent reproduction, in the case where the encoded image to which the order is added is the reference image B to which reference should be made when another encoded image is decoded.

Embodiment 4. The apparatus for generating information trains with the characteristics of the form of embodiment 3, wherein said addition unit is operative to add the repetition information to one between an image I and an image P that follows, in order of decoding, the encoded image that is valued to be skipped, being the encoded image a reference image B.

Embodiment 5. The apparatus for generating information trains with the characteristics of the form of embodiment 1, wherein said valuation unit is operative to assess that an image P is skipped at the time of the intelligent reproduction, in the case where the encoded image to which the order is added is the image P that must skip when a specific P image is decoded, and the specific image P can be decoded by selective decoding of an image I or image P precedent in decoding order.

Embodiment 6. The apparatus for generating information trains with the characteristics of the form of embodiment 5, wherein said addition unit is operative to add the repetition information to the other image that follows, in order of decoding, to the image P that is valued to be skipped and that is necessary to decode the specific P image.

Embodiment 7. The apparatus for generating information trains with the characteristics of the form of realization 6, wherein said valuation unit is also operative to assess that a reference image B must skip at the time of smart playback, in the case where the encoded image to which the Order is the reference image B.

Embodiment 8. The apparatus for generating information trains with the characteristics of the form of realization 7, wherein said addition unit is further operative to add the repetition information to one between a image I and an image P that follows, in order of decoding, the reference image B that is valued that it should skip over.

Embodiment 9. A procedure for generating information trains to generate a train of information that includes encoded images and an order to manage a buffer that saves a decoded image as a reference image, adding the order to one of the encoded images that are used as a reference image, said procedure comprising: assessment of whether the encoded image to which the order is added should be skipped or not at the time of the smart reproduction; addition, in the case where it is assessed that the encoded image must be skipped, of repetition information indicating the same content as the order to another encoded image that follows, in decoding order, the image coded that is valued that should be skipped and that is not skipped at the time of intelligent reproduction; Y Generation of the information stream that includes coded images, order and repeat information.

Embodiment 10. An image coding apparatus comprising: an operational coding unit for encoding images; a first operational addition unit to add an order to manage a buffer that stores a decoded image as a reference image to one of the encoded images used as the image of reference; an operational assessment unit to assess whether the encoded image to which the order is added should be skipped or not at the time of intelligent reproduction; an operative addition unit to add, in the case where it is assessed that the encoded image must be skipped, Repeat information indicating the same content as the order to another encoded image that follows, in order decoding, to the encoded image that is valued that should be skipped and that is not skipped at the time of smart reproduction, and an operational generation unit to generate the information stream that includes the encoded images, the Order and repeat information.

Embodiment 11. An image coding procedure, comprising: image coding; adding an order to manage a buffer that saves a decoded image as an image of reference to one of the encoded images that is used as the reference image; assessment of whether the encoded image to which the order is added should be skipped or not at the time of the smart reproduction; addition, in the case where it is assessed that the encoded image must be skipped, of repetition information indicating the same content as the order to another encoded image that follows, in decoding order, the image coded that is valued that should be skipped and that is not skipped at the time of intelligent reproduction; Y Generation of the information stream that includes coded images, order and repeat information.

Embodiment 12. A computer-readable recording medium in which a program is recorded, using the program to make a computer execute the procedure of generating information trains with the characteristics of the embodiment 9.

Embodiment 13. A computer readable recording medium in which a program is recorded, using the program to make a computer execute the image coding procedure with the characteristics of the embodiment 11.

Embodiment 14. A program to make a computer execute the procedure for generating Information trains with the characteristics of the embodiment 9.

Embodiment 15. A program to make a computer execute the coding procedure of images with the characteristics of the embodiment 11.

Embodiment 16. An information stream that includes encoded images, wherein said information train includes a repeat order and information, said order is an order to manage a buffer that saves a decoded image as an image reference, and is added to one of the encoded images used as the reference image, and said repetition information has the same content as the order added to one of the encoded images which should be skipped at the time of smart playback, and added to another encoded image that follows, in decoding order, to the encoded image that must be skipped and not skipped at the time of smart reproduction

Embodiment 17. A computer readable recording medium in which the information stream is recorded with the characteristics of the embodiment 16.

Embodiment 18. An integrated circuit that generates an information stream that includes encoded images and an order to manage a buffer memory that saves a decoded image as a reference image, adding the order to one of the encoded images, said integrated circuit comprising: an operational assessment unit to assess whether the encoded image to which the order is added should be skipped or not at the time of intelligent reproduction; an operative addition unit to add, in the case where it is assessed that the encoded image must be skipped, Repeat information indicating the same content as the order to another encoded image that follows, in order decoding, to the encoded image that is valued that should be skipped and that is not skipped at the time of smart reproduction; Y an operational generation unit to generate the information stream that includes the encoded images, the Order and repeat information.

Claims (3)

1. An information train generation apparatus that generates an information stream for intelligent reproduction, according to MPEG-4 AVC, which includes (i) OCGM memory management control operation order to manage a buffer memory (2004) that saves a decoded image as a reference image, and (ii) encoded images, said apparatus comprising: a generation unit (1004) operative to generate the encoded images by image coding, an operational order addition unit for adding the OCGM to a predetermined encoded image, characterized by: an evaluation unit (1002) operative to assess whether a first encoded image to which OCGM is added is or is not a reference image B that can be referenced when another encoded image is decoded; Y a repeating unit (1003) operative to generate a supplementary reinforcement information of the decoded reference image markup ISR RMIRD when said valuation unit (1002) assesses that the first encoded image is the reference image B and is added with OCGM; wherein said generating unit is operative to generate the information stream by adding the RMIRD ISR to a second encoded image I or a second encoded image P, so that said second image follows the first encoded image in decoding order. 2. An information train generation procedure for generating an information stream for intelligent reproduction, according to MPEG-4 AVC, which includes (i) OCGM memory management control operation order to manage a buffer memory (2004) that saves a decoded image as a reference image, and (ii) encoded images generated by image coding, in which the OCGM is added to a predetermined encoded image, with said procedure characterized by the steps of: assess (S1002, S1102) whether a first encoded image to which first information is added is or is not a reference image B to which reference should be made when another encoded image is decoded; generate (S1003) supplementary reinforcement information for marking repetition of a decoded reference image ISR RMIRD, repeat OCGM by entering OCGM data to the RMIRD ISR, when it is valued in said valuation that the first encoded image is the reference image B and added with OCGM; Y generate (S1008) the information stream by adding ISR RMIRD to a second encoded image I or a second encoded image P, so that the second image follows the first encoded image in decoding order. 3. A computer readable recording medium in which an information stream for intelligent reproduction is recorded, according to MPEG-4 AVC, the information stream being generated by means of a method as defined in claim 2.