MXPA96003031A - Method and data coding / decoding device and data logging means codifica - Google Patents

Abstract

The present invention relates to an apparatus and method for recording encoded image data in a recording medium that operates to receive the image data (e.g., a video data), encode the image data using intraframe coding and / or predicative coding to provide an I image and two successive P images, generate position information representing the positions of the two I images, two P images, and the position information in the recording medium. The encoded image data is reproduced from the recording medium in a special reproduction mode by selective reading of the data using the position information.

Description

METHOD AND DEVICE FOR ENCODING / DECODING DATA AND MEANS OF REGISTERING CODED DATA BACKGROUND OF THE INVENTION The present invention relates to a method and apparatus for encoding data that can encode audio and video data for special reproduction, to a method and apparatus for data declassification to read video and audio data recorded from an optical disk, a magnetic disc or the like and reproduce the data read in a special way, and refers to a record medium where the encoded data is recorded in such a way that they can be played in a special mode. Digital or similar image signals to be recorded on a disc in a digital video disc system < below DVD) are compressed and encoded using the MPE8 method < 8 group of Experts of Image Coding in Motion). Fig. 14A is a schematic representation of the interquarter prediction structure employed in the MPEG system. In this example, a GOP (Group of Images) consists of, for example, fifteen frames, which include a frame of an I image (interframe encoded image), four frames of P images (images encoded by direct prediction of ipterframe), and the remaining frames consist of B images (images encoded by bidirectional, direct and indirect prediction). Image I is an interframe encoded image in which a frame is compressed data; the image P is an image encoded in a direct interframe predictive form in which a frame is encoded by compression with reference to the preceding frame (picture I or picture P) already encoded; and image B is an image bidirectionally bidirectionally encoded in which a frame is encoded by compression with reference to temporarily before and after frames. More specifically, as indicated by the arrows in the diagram, an I, 10 image is encoded by means of interframe processed by itself and without reference to any other frame, an image P, PO, is encoded by prediction interframe with reference to the image I, 10; and an image P, Pl, is coded by interquarter prediction with reference to the image P, PO. In addition, the images B, BO and Bl are coded by interframe prediction with reference to both the image I, 10 as well as the image P, P0; and the images B, B2 and B3, are coded by interframe prediction with reference to both the image P, P0 camo to the image P, Pl. In the same way, subsequent images are encoded by such prediction in the manner indicated by the arrows .

When decoding the images encoded in predicted form a, the I image is decoded alone since na is encoded with reference to no other frame. However, a previous image I or a previous image P refers to decoding a given image P because an image P is encoded in a predictive manner with reference to the image I or the image P temporarily prior. In the same way, prior or subsequent I images or P images are referred to decode a given B image because a B image is encoded with reference to the I images or to the images P temporarily, preceding or subsequent. For this reason, in order to provide an adequate decoding, the positions of the images in a recording medium of FIG. 14A to the positions illustrated in FIG. 14B in such a way that the images required for declassification are decoded in advance. As illustrated in the diagram, such position changes are made in such a way that the image I, 10, precedes the images B, Bl and B-2, since the images I, 10, are required to declassify the images B, Bl and B-2, and also the position of the image P, PO, is changed to precede the images B, BO and Bl, since the descadif ication of the images B, BO and Bl, requires the image I, 10, and of the image P, P0.

In the same way, other images are changed in such a way that the image P, Pl precedes the images B, B2 and B3, since the images P, PO and Pl, are required to descadify the images B, B2 and B3, and also the image P, P2, precedes the images B, B4 and B5, since the desei fi cation of the images B, B4 and B5, require the images P, Pl and P2. In the same way, position changes are made in such a way that the image P, P3, precedes images B, B6 and B7. The composite video data of the image I, images P, and images B, arranged in the order of fig. 14B, and other data including audio data and subtitle data, are polled and recorded in the recording media such as a disc or transmitted in a transmission channel. The amount of code of each frame in the image data is set between images and depends on the complexity or simplicity of the individual image. Typically, an I image is represented by a greater number of data in an image P that is represented by a greater number of data than an image B. Figs. 15A to 15C show an example of how the data can be determined iplexan. In these diagrams, fig. 15A represents an MPEG2 system stream that is ultiplexed after its packet formation; fig. 15B represents the content of a video packet in the multiplexed stream; and fig. 15C represents an MPEG2 video stream of a video layer. In each of the image data V, V + l, V + 2, ... etc. which constitute the video layer of a fig. 15C, the image header information and the image coding extension information is placed in the initial position. In the example shown, a video stream that is located from the position identified as DI to the identified position cam D3 of the video layer forms a video package with a packet header placed in its initial position and a video stream that is located from position D3 to the position identified as D5 of the video layer forms another video packet with a packet header placed in its initial position. The video packets are multiplexed with audio packets and subtitle packets to form the MPEG2 system stream presented in FIG. 15A. Fig. 16 shows the contents of an image header, and fig. 17 shows the contents of an image coding extension. In the image header, there is information such as a unique image start code, a temporary reference (TR) (which is a temporary serial number provided by image), and a type of image coding (image I, P or B). In the image coding extension there are information elements such as a unique extension start code, a unique extension start code identifier, an image structure, a first upper field, a progressive frame, etc. As for the image data, two data structures can coexist: a frame structure where an image consists of a frame and a field structure where an image consists of two fields. If the image data has a frame structure of one frame per image or a field structure of two fields per image it can be identified from the following three elements of information, namely (1) presence of a GOP header, (2) ) temporal reference (TR) in the image header, and (3) image structure in the image coding extension. Fig. 18 is a block diagram illustrating an example of the data decoding apparatus adapted to perform a special reproduction of data such as a slow reading d image, a fast reading, a reverse reading and the like. An optical disk 1 rotates by means of an axis motor (not shown) at a predetermined rotation speed, and a laser beam is projected from a reader 2 towards a track of the optical disk 1 in such a way that the digital data MPEG tablets registered in the track can be read. The digital data is processed by means of a demodulation circuit 3 that demodulates from eight to fourteen modulations (EFM) and supplied to a sector detection circuit 4. The output of the reader 2 also has a phase-locked circuit 9 (PLL), where a clock signal is reproduced and supplied to the demodulator circuit 3 and to the sector 4 detection circuit. The digital data recorded in the disk 1 include lexical groups registered in units of a sector of full length, with a syn- onization of sector and a heading of sector placed at the site of each sector. The sector detection circuit 4 detects each of the sectors from the sector direction and sector synchronization of the sector header; and this information is supplied to a control circuit 6. The demodulated digital data is supplied, by means of the sector detection circuit 4, to an EEC circuit (error correction) 33 which executes the detection and correction of errors. The ECC circuit 33 provides corrected data to an annular separator 5 so that they are written to the under control of control circuit 6. The output of the ECC circuit 33 is also supplied to a flow detector 50 which determines the type of image from the data stream image header read from disk 1 in special playback mode and then supplying image type information to control circuit 6. In response to this information, control circuit 6 executes its control operation in such a manner that, in the special operation mode, the data of the image I and 5 the data of the two subsequent P images are written to the annular separator 5. A focus control circuit (not shown) and a servo tracking circuit 8 control the focus and tracking of the reader 2, respectively, under the control of a controller system (not shown) in response to a focus error signal and a tracking error signal obtained from the information read by the reader 2. According to the sector address of each sector detected by the detection circuit 4 sector, the control circuit 6 designates, by means of a write indicator WP, a write address for wri the corresponding sector in the annular separator 5. Furthermore, according to a coding request signal obtained from a video code regulator 10 (Fig. 1SB ), the circuit The control 6 also designates, by means of a reading indicator RP, a reading direction of the data written on the annular separator 5. The control circuit 6 is adapted to read the data from the position of the reading indicator RP and supplies the read data to a tiplexor demult - fts Since the encoded data recorded on the disk 1 comprises subtitled data, audio, multipared video, the desmu.lt iplexor 32 separates the data read from the annular separator 5 into video data, audio data and subtitle data, and then supplies the respective data to a video decoder 20 (Fig. 18B), an audio decoder (not shown), and a subtitle decoder (not shown). The video decoder 20 stores the video data in the video code separator 10. Then, the data stored in the video code separator 10 is supplied to an image header detector 34 which detects the header of the image. . The image header information detected is additionally used to identify the type of image (image I, P or B) of the video data and the temporal reference (TR) that indicates the order of the frame in the GOP. An image data selection circuit 35 selects only the image I and the image P as identified by the image type information supplied from the image detector 34 in special playback mode, and supplies the data of the selected image to an 11 reverse VLC circuit (variable length coding). In a normal reproduction mode, the image selection circuit 35 is controlled to supply all the image data to an inverse VLC circuit 11 without any preselection. The data supplied to the inverse VLC circuit 11 is processed using inverse VLC; and then supplied to a decoder 12. The code request signals are returned to the video code controller 10 from the inverse VLC circuit to allow new data to be transferred from the video code separator 10. Likewise, the inverse VLC circuit 11 emits a quantization step size to the dequantizer 12 and outputs a motion vector information to a motion compensator 15. The quantization step size information and the motion vector information they are included with the video data. The discoverer 12 discovers the data entered according to the size of the designated quantization step and outputs the dequantized data to a reverse 13 DCT (discrete casinusoid l tr) circuit. The inverse DCT circuit 13 processes the dequantized data using reverse DCT to retrieve video information, and supplies the recovered video information to an adder 14. The adder 14 adds the output of the inverse DCT circuit 13 and the output of the motion compensator 15 according to the type of image (I, P or B) and supplies the result, ie the video data compensated by movement, to a bank 16 of frame memory.

Then, the data read from the block memory bank 16 is rearranged in the original block order (as shown in Fig. 14A by means of the switch 16E). The reordered data is supplied to a digital-ana logical converter (D / A) that transforms the data into an analog video signal for viewing in a viewing device 18. Returning to fig. 18, the output of the E-CC circuit 33 is supplied to a flow detector 50 which detects the type of image from the flow data read from the disk 1 and supplies the image type information to the control circuit 6. In response to this information, the control circuit 6, in the special reproduction mode, writes in the annular separator 5, the data of the image I and the two subsequent images P. Accordingly, three frames corresponding to the image I and the images P are written at the beginning of each GOP at high speed in the annular separator 5, and these data can be acquired and decoded by means of decoder 20 at the desired time, thus allowing an efficient decoding of the data in the special reproduction mode. Suppose, for example, that a reverse reproduction is started with the image P, P3 of the original order of pictures presented in fig. 14 TO. It is necessary to show the images decoded in the following order: P -B7-B -P2-B5-B4-P11-B3- 92-po-Bl-BO-10- ... However, since each image P is encoded by interframe prediction, as described, the images 10, PO, Pl and P2 require their decoding before the descaling of the image P, P3. -In the same way, the images P, P2 and P3 require their decoding before the decoding of the image B, B7. Accordingly, if an inverse reproduction is performed by decoding each image only once, as in normal reproduction, it is necessary to employ a high capacity frame memory bank 16 which can store the same number of frames as the number of images that constitute a GOP. The storage capacity of a frame memory bank 16 should be increased beyond what is required in a normal reproduction mode to satisfy such requirements. Likewise, the decoded data must be stored sequentially in the frame memory bank in order to supply the images in the proper order of reverse reproduction. Even though other reverse reproduction techniques can be adopted to perform only the I and P images, thus skipping the B images, the need to store more frames than is required for normal reproduction continues to exist. For this reason, the data decoding device of FIG. 18 operates to perform a reverse reproduction using the same frame memory bank as the one used in normal reproduction, that is, using three memory elements in the example of fig.18 to store an I image and two succeeding P images temporarily. The flow detector 50 installed for this purpose writes the image I and the two subsequent images P to the annular separator 5. However, this causes the construction and operation of the flow detector 50 to detect the image I and the two images P subsequent ones are more complicated. Fig. 15A shows the system flow MPE62 in the form of packets (many iplexed). When an MPEG2 video stream packet is defined at a position D2 as shown in FIG. 15C during the packet formation process, the image header and the image coding extension of the image data (V + 2) are spread over two packets, as shown in FIG. 15B. If the image header and the image coding extension are spread over two video packets, it is necessary to detect two video packets to obtain the necessary information elements of the image. In addition, as shown in fig. 15A, another package (e.g., audio package) may exist between two video packets to complicate the detection process, thus complicating the construction and operation of the flow detector 50. According to the MPEG2 technique, the video data in a structure of pictures where an image is composed of a picture, and video data in a field structure, where an image consists of two fields can be intermixed. Since an image header is placed over each field, image headers and image coding extensions of two consecutive images must be read to determine the data structure of the video data. Accordingly, a determination is made for make sure if the image data are formed in the appropriate structure or in the field structure based on the three aforementioned information elements, that is, (1) presence of a GOP header; (2) temporal reference (TR) in the header of the image; and (3) image structure information in the image coding extension. Next, a detailed explanation of a method of differentiation between a frame structure and a field structure will be presented. Figs. 19A and 19B show video data in frame structure formats, respectively. In the field structure format, a video data frame is composed of two fields of image data, each with an image header and an image coding extension. In the frame structure format, a video data frame is composed of an image data frame to which an image header and an image coding extension are fixed. In the field structure format, the numerical values of the information TR in the respective image headers of the image data pairs are determined to be equal to each other. The image structure information in the image coding extension is "01" and "10" for the Upper and Lower Field, respectively, as shown in FIG. 20. In addition, the information of the image structure in the image coding extension of the frame structure is "11" as shown in FIG. 20. The format (field or frame structure) of the image data should be checked by a first reading of the GOP header at the initial GOP position and then by reading the image structure information of the coding extension of GOP. image at the beginning of the data in this image. Although the image data in the frame structure can be loaded into the annular separator 5 (Fig. 18A) by detecting a single frame, it is difficult to load the video data in the same way in the field structure where a pair of image data constitute a video data frame, because the image data in pairs must be detected before they can be properly loaded. Accordingly, the information TR in each image header is read to find two units of image data with numerically equal values in TR. When such a pair is found, they are identified as image data in pairs and loaded. The field structure image headings in pairs are arranged in two different possible orders: up / down and ab / up. Such arrangements will now be described with reference to fig. 21. A GOP header (GOP H), a quad frame I image, a field structure B image, another field structure B image, a spaced GOP header, and a field structure I image , a GOP header, another I image of field structure ... etc. they are registered sequentially. For example, where a total of three frames (one I image and two subsequent P images) is loaded into an annular separator 5 (FIG. 18A), the I image in the frame structure that follows the upper 60P header (first) is detects and identifies from the GOP header, the type of image coding in the image header, and the image structure information ("11" in the case of the frame structure), in the image coding extension at the beginning of the image datas. When reading is accessed in the position identified as random access 1 in the byte stream, the image header as well as the image coding extension of the first frame B of the field structure are read. At this time, TR is also read as "0". Subsequently, the image header and the image coding extension of the second field structure B image are read as well as TR expressed as "0". Since the TR value of two B-images of field structure are equal, it is detected as data in pairs. When two bytes in the byte stream are accessed at the position identified as random access, the image header and the image condion extension of the first image are read together with the TR expressed as co or "". Subsequently, the image header and the image coding extension of the following frame are read together with the TR expressed as "1". Since the respective numerical values of the TR do not coincide with each other, the data of the two field structure boxes are not detected with data in pairs.

If you have access to the position identified as three random access in the byte stream, the respective numerical values of the TR in the two image headers coincide (TR = 1) between them co or in the previous case of random access 1, allowing thus the detection of the images as a pair. If an image structure is detected in the image coding extension of "01" or "10", a field structure is considered camo, and data in pixels is detected. If one has access to the identified position co or the random access 4 in the byte stream, the image header and the image coding extension of the first image are read together with the TR expressed as "0". These image datas are considered as an I field structure image in accordance with the image coding type information in the image header and the image structure information in the image coding extension. After the subsequent detection of the GOP header, the image header and the image coding extension of the next image are read together with the TR expressed as "0". Here, the respective numerical values of the TR of two consecutive images coincide with each other, but these two images are not considered as paired because there is a GOP header between the two images. It should be noted that the TR is set to "0" if there is a GOP header and there is no GOP header interposed between a pair image. In accordance with that described, the flow detector 50 performs a process of detecting various elements of information relating to the images corresponding to the GOP header, the image headers, and several flags of the image coding extensions, to load the image data in the annular separator 5. However, this processing routine is extremely complex making it difficult to construct the flow detector 50. The way in which the flow detector 50 detects the end of the charging process is explained in combination with the flow diagram presented in figure 22. It is considered in this flowchart that random access to an input sector written immediately prior to an I image is performed in such a way that an adequate image can be obtained instantaneously in response to random access. In the SIO step, the flow detector searches for the image start code in the image header to detect the image header of the image I and an inquiry is made in step S12 to determine whether the image start code is He has detected. If the inquiry is answered in the affirmative in step S12, that is to say -o if the image start code is detected, the operation continues with step S14. However, if the inquiry of step S12 is answered in a negative way, that is, if no image start code is detected, the process of step S12 is repeated until the image start code is detected. In step S14, the time reference is read from the detected image header and its numeric value is stored in a register as TRO. In step S16, another search (SRCH) of the image start code in the image header is performed to detect the next image and a search is performed in step S18 to (determine if the image start code has been detected. If the inquiry of step S18 is answered in the affirmative, that is, if the image start code is detected, the operation proceeds to step S20, however, if the response to the inquiry of step S18 is negative, it is repeated the process of step S18 until the image start code is detected In step S20 an inquiry is made to determine if a GOP header has been detected in the image start code, so that the data is determined Detected images are part of a pair If the answer to S20 is answered in a negative way, that is, if no GOP header is detected, the operation continues *? 1until step S22. However, if this inquiry is answered in the affirmative, that is, if the SOP header is detected, the operation proceeds to step S26 because the existence of a GOP header between the image data eliminates the possibility that these units will image were paired. When the time reference is read from the detected image header, its numerical value is stored in a register as TRI as represented in step S22, and the operation proceeds to the question in step S24 to determine whether the numerical values of the TR stored respectively in the register TRO and TRl are igu les. If the question in step S24 is answered affirmatively, that is, if there is a match between the two numerical values, the operation returns to step S16 and the processes outlined above in relation to steps S16 to S24 are repeated. It will be noted that the coincidence of two numerical values means that a pair of image data has been detected. However, if the question of step S24 is answered in a negative way, ie if the numerical values of the TR are not equal, the operation continues until step S26. Here, the image header of the next image is detected; and the type of image code read from the image header is stored in the register. The operation then proceeds to the question of step S28 to determine whether the type of stored image coding represents an image B. If the question of step S28 is answered in the affirmative, that is, if the detected image is a B image, the operation return to step S16 because the image B na is being searched; and the processes presented above in relation to steps S16 to S28 are repeated to detect the next image. However, if the question of step S28 is answered in a negative way, ie if the detected image is not a B-image, the time reference in the detected image header is read and its numerical value is stored in the comma-record TR2, as represented in step S30. It will be noted that this image-detected is the first image-P that appears after the image I. In step S32, another search (SRCH) is performed for the image start code in the image header to detect the next image and in step S34 a question is posed to determine if the image start code has been detected. If the question in step S34 is answered in the affirmative, the operation proceeds to step S36. However, if the question of step S34 is answered in a negative way, that is, if no image start code has been detected, the process of step S34 is repeated until the image start code is detected. In step S36 a question is asked to determine if a GOP header has been detected during the search of the image start code, to determine whether the detected image data is part of a pair. If the question in step S36 is answered in a negative way, ie if no GOP header is detected, the operation continues until step S38. However, if the question in step S36 is answered in the affirmative, that is, if a GOP header is detected, the operation proceeds to step S42 because the existence of a GOP header between the image units eliminates the possibility that these imaging units are paired. When the time reference is read from the image header detected in step S38, its numerical value is stored in the register as TR3 and the operation proceeds to the question in step S40 to determine if a match is achieved between the numerical values of TR stored respectively in the register as TR2 and TR3. If the question of step S40 is answered in the affirmative, that is, if the two numerical values are equal, the operation returns to step S32 and the processes presented above in relation to the passes S32 to S40 are repeated. It will be noted that the coincidence of two numerical values means that a pair of image data has been detected.

However, if the question of step S40 is answered in a negative way, that is, if the numerical values of TR na coincide, the operation continues until step S42 to read the image type. The operation then continues until the question of step S44 to determine whether the type of stored image coding represents an image B. If the question of step S44 is answered in the affirmative, that is, if the detected image is a B image, the operation returns to step S32 because the image B was not looking for; and the processes outlined above in relation to steps S34 to S44 are repeated to detect the next image. However, if the question of step S44 is answered in a negative way, ie if the detected image is not a B-image, the 3 rd time reference detected in the image header is read and its numerical value is stored in the register, TR4, as it is represented in step S46. It will be noted that this detected image is the second image P that appears after the image I. Following with step S48, another search (SRCH) is made for the image start code in the image header to detect the next image and In SSO a question is asked to determine? I have detected the image start code. If the question of step S5 is answered in the affirmative, that is, if the image start code is detected, the operation proceeds to step S52. However, if the question of step S50 is answered in a negative way, the process of step S50 is repeated until the image start code is detected. In step S52, a question is asked to determine if a GOP header has been detected during a search of the image start code, determining whether the detected image is part of a pair. If the question of the SSO step is answered in a negative way, that is, if no GOP header is detected, the operation continues until step S54. However, if the question of step S52 is answered in the affirmative, that is, if a GOP header is detected, the loading of the image data in the annular separator is completed and the process is completed. When the time reference is read from the detected image header and its numerical value is stored in the co or TR5 register, as shown in step S54, the operation proceeds to the question in step S56 to determine whether agreement is achieved between the numerical values of TR stored respectively as TR4 and R5. If the question of step S56 is answered affirmatively, that is, if the two numerical values coincide, the operation returns to step S48 and the processes raised above in relation to steps S48 to S56 are repeated. However, if the question in step S56 is answered in a negative way, that is, if the two numerical values of TR. they are not the same, the loading of the image data is complete and the process ends. Accordingly, the flow detector 50 can load a bit stream by executing the above processing routine to load an I image and two successive P images. However, as can be seen from this long description, it is quite expensive to execute this complex processing routine. OBJECTS OF THE INVENTION Accordingly, an object of the present invention is to provide method and apparatus for encoding data to perform a special reproduction, for example, fast forward and backward reproduction, which overcomes the limitations of the technique described above, ie , the complicated operations mentioned above. Another object of the present invention is to provide a method and apparatus for de-dataing to perform special reproduction that overcomes the limitations of the technique described above. A further object of the present invention is to provide a recording means for use in combination with a processor-controlled apparatus for performing a special reproduction without requiring the processor-controlled apparatus to perform the abovementioned complex operations. Various other objects, advantages and features of the present invention will be readily apparent on the basis of the following detailed description and the novel features will be especially indicated in the appended claims. SUMMARY OF THE INVENTION In accordance with one embodiment of the present invention, apparatus and method are provided for recording encoded image data in a recording medium. The image data is encoded by the use of intraframe coding and / or predictive coding to provide an I image and two successive P images. Position information is generated which represents the positions of two images P in relation to the image I, and the image I, the two images P, and the position information are recorded in the record mean. As an aspect of the present invention, the position information represents the length of data in bytes from the image I to the end of the first image P and / or to the end of the second image p. According to another embodiment of the present invention, apparatus and method for reproducing image data encoded from a recording medium are provided. The position information representing the positions of two P images relative to an I image is detected, and a data stream including the I image, the two P images, and the position information is generated. The data flow is decoded and viewed. In accordance with another embodiment of the present invention, a recording means is provided for its use and combination with a processor controlled apparatus where an I image is recorded, two P images, thus bed position information representing the positions of the two P images in relation to the image I, on the medium and is used by the processor-controlled apparatus to perform the special reproduction relatively easily and simply. BRIEF DESCRIPTION OF THE DRAWINGS The following detailed description, provided as an example, is not intended to limit the present invention only to what is presented herein, and would be better understood in connection with the accompanying drawings in which: Fig. 1 is a block diagram illustrating one embodiment of the data coding apparatus of the present invention; The. Fig. 2 shows an example of a packet-shaped stream encoded by the data coding apparatus of fig. 1; Fig. 3 shows a presentation of information of the entry point; Fig. 4 shows a syntax of a program flow map (SPM); Fig. 5 shows a syntax of elementary flow descriptors. Fig. 6 shows a syntax of an ip ipp descriptor; Fig. 7 shows a syntax of global descriptors; Figs. 8A and 8B are a block diagram illustrating one embodiment of the data decoding apparatus of the present invention; Figs. 9A to 9C are diagrams showing an example of the order of video data to which reference will be made in describing how the video data is read in a fast reverse reproduction mode of the data decoding apparatus of Figs. 8A and 8B; Fig. 10 is a read / write synchronization diagram which will be referred to when describing the fast reverse reproduction mode of the data decoding apparatus of Figs. 8A and 8B; Figs. HA at 11C are diagrams showing an example of the order of video data which will be referenced in describing how video data is read in a fast forward playback mode of the data de-referencing apparatus of FIGS. 8A and 8B; Figs. 12A to 12C are diagrams showing another example of the order of video data which will be referred to when describing how the video data is read in a fast reverse reproduction mode of the data decoding apparatus of FIGS. 8A and 8B; Fig. 13 is a read / scan synchronization diagram which will be referred to when describing how video data is read using two frame memories in the fast reverse playback mode of the data decoding apparatus of Figs. 8a and 8B; Figs. 14A and 14B are schematic representations of structures of the original interframe prediction images and image registration frames in the MPEG system; Figs. 15A to 15C are schematic representations of an MPEG video stream; Fig. 16 shows a structure of an image header in the MPEG system; Fig. 17 shows a structure of an image coding extension in the MPEG system; Figs. 18A and 18B are a block diagram illustrating a data decoding apparatus; Figs. 19A and 19B are schematic representations of the video data structure in a frame format and in a field format; Fig. 20 is a table illustrating the content of an image structure; Fig. 21 is a diagram of a video stream showing how two video formats (field and frame) differ; and Fig. 22 is a flow diagram of a routine executed in a flow detector to load three frames of the image data (one I image and two P images) where the video, field and frame formats are in. remixed DESCRIPTION OF THE PREFERRED MODALITIES Preferred embodiments of the present invention are described below in detail with reference to the appended drawings. Fig. 1 is a block diagram showing one embodiment of the data coding apparatus of the present invention. An audio encoder 102 encodes by compressing an input audio signal supplied to it. The audio input terminal and a video encoder 101 encode by compression an input video signal supplied to its video input terminal. The encoded audio and video signals are supplied to the multiplexer 113. It is considered that a stream left from the audio encoder 102 is an MPEG2 audio stream (audio layer) and a stream left from the video encoder 101 is an MPEG2 video stream. (video layer), the latter being represented in FIG. 15C. The multiplexer 113 ultiplexes the input MPEG2 video stream and the MPEG2 audio stream by time division multiplexing to form the system stream, presented in FIG. ISA. Although not shown, the subtitle stream can also be entered into the multiplexer 113 and can be multiplexed with the video stream and the audio stream. In such a case, the MPEG2 system stream left from the multiplexer 113 is as shown in FIG. ISA An input terminal of an input point data memory circuit 133A is connected to the video encoder 101; and an entry point detector 131 allows the input point data memory circuit 133A to store an input point (data related to an image generation point I) received from the video encoder 101. A data generator 156 Table of Contents (TOC) generates TQC data based on the contents of the 133A circuit of the input tip data memory. The TOC data includes a name of the disc - on which the audio and video data are recorded, the name of each chapter registered on the disc, the initial address of each chapter on the disc, the playing time of the disc, the time of reproduction of each chapter, the initial address of each entry sector and the like. The multiplexed stream leaving the multiplexer 113 is stored temporarily in a DSM (Digital Storage Medium) 110 and then administered to a suffix circuit 150 TOC. The TOC suffix circuit 150 adds the TOC data to the multiplexed stream and supplies said stream to an image header detector / generation device - override of program flow map data (PSM) 155. The image header / generation device - PSM data override 155 detects an image header that generates PSM data including the presentation of data length information in bytes from the beginning of an input sector to the end of the first image P that appears and / or from the beginning of the input sector until the end of the second image P that appears "e." PSM data preferably includes information representing the length of data in bytes from the beginning of an I image to the end of the first P image that appears and / or from the beginning of an I image to the end of the second P image that appears.The generated PSM datas are written to an area in the previously reserved input sector. for this in the stream multiplexed by the multiplexer 113. A detailed description of the PSM data will be presented below.

The output of the image header / generation device - PSM data sabreration device 155 is supplied to a sector header suffix circuit 151, where the multiplexed stream is divided into sectors with the addition of a sector header to each of the sectors. The output of the sector header suffix circuit 151, i.e., the video and audio data together with all the other aggregate data, as described above, is encoded for error correction by means of an ECC 152 encoder. a modulator 153 modulates the data encoded by the ECC encoder 152 using 8 to 14 modulations (EFM) and the modulated data is supplied to a cutting machine 154. The cutting machine forms depressions in a master disk 160 according to the data supplied to starting from the modulator 153, so that the flow data are typed iplexed on a 160 Master DVD disk. A copy of the DVD disc is produced, for example, by the pressure molding of this master disk 160. Accordingly, the data coding apparatus of FIG. 1 encodes and ultiplexes, using time division multiplexing, an audio signal and an input video signal to produce a multiplexed stream. In addition, the image header / generation device - PSM data saber 155 is generated and writes PSM data in the multiplexed stream. The ultiplexada flow registers in the disc 160 DVD master. Fig. 2 shows an example of the multiplexed stream, for example an MPEG2 system stream, leaving the image header detector / generation device - PSM 155 data overwriting. For simplicity, only audio and video data are shown. imitated video data. The audio data is inserted into certain parts of the MPEG2 system stream to ensure that the sound is not interrupted during playback, and the video data of the I, P and B images are inserted between the audio data. Entry point means the upper (or initial) position of an I image, and a sector that includes such an entry point is called the input sector. In fig. 2, the positions of such entry points are indicated as entry point n, entry point n + l, ... etc. The position in which the entry point information is written is predetermined to be immediately before an I image, such that an entire image can be instantly viewed when a reader has read the input sector data. There may be audio data between the input point information and an I image, but there are no images P and B 56 between them. Fig. 3 shows a presentation of entry point information. The entry point information includes a packet header with an optional system header, a PSD (Program Flow Directory), a PSM (Program Flow Map) and other packages. Fig. 1.4 shows the PSM syntax. The PSM includes a 24-bit packet start code prefix that forms a unique code, an 8-bit map flow identity, a program flow information composed of an arbitrary number of global descriptors, a flow type, and an elementary flow information that includes an arbitrary number of elementary flow descriptors. Fig. 5 shows a syntax of the elementary stream descriptors composed of DVD video descriptor and ip ipp descriptor if the stream is video data, to DVD audio descriptor composite and ISO 639 language descriptor if the stream is data from audio, or composed of DVD subtitle descriptor and ISO 639 language descriptor if the stream is subtitle data. Likewise, other elements of information are presented in fig. 5. As shown in Fig. 6, the ip ipp descriptor consists of an 8-bit descriptor tag which means an ip ipp descriptor, a description length of 8 bits which means the length of the descriptor, bytes up to -7 the first 32-bit P disk which means the number of bytes from the first byte of the present input sector to the last byte of a P image that first appeared, and bytes to the second 32-bit P image which means the number of bytes bytes from the first byte of the present input sector until the last byte of the image P that appears second. The bytes up to the first image P and the bytes up to the second image P represent the data lengths, as shown in fig. 2. It is noted that the number of compensated bytes indicated by the information of bytes to the first image P and of bytes to the second image P includes not only I and P images but also intermediate images P and audio packets, as shown in FIG. fig. 2. Fig. 7 shows a syntax of the global descriptors of fig. 4. The PSD included in each input sector represents the distances between the present input sector and the previous input sector and the next input sector, and the distances between the input sectors after a time of one second, three seconds , etc. These distances are known as deviation directions. Figs. 8A and 8B are a block diagram of a preferred embodiment of the data decoding apparatus of the. present invention. For simplicity, the elements specified in fig. 8A and 8B corresponding to the elements presented in FIGS. 18A and 18B are indicated by means of the same reference numbers. An optical disk 1 rotates at a predetermined rotation speed by means of the action of an axis motor (na shown), and a laser is projected from a reader 2 towards a track of an optical disk 1, in such a manner that the digital MPEG data recorded in the ista is read from it. The digital data is demodulated EFM by means of a debunking module 3 and sent to a sector 4 detector. The output of the reader 2 is also supplied to a phase 9 synchronization circuit (PLL) which requires a signal of no snoring. n that is supplied to both the demodulator 3 and the sector 4 detector. As previously mentioned, the digital data is recorded on disk 1 in units of fixed length sectors, where a sector synchronization and a sector header are recorded at principle of each sector. The division of sectors is determined from the detection of the sector synchronization and the sector address from the sector header that is supplied to a control circuit 6. Preferably, the control circuit is implemented by means of micropracesadares and performs the control by processors of the illustrated apparatus.

The demodulated output is supplied via the sector detector 4 to an ECC circuit (error correction) 33 which executes error detection and correction. The ECC circuit 33 supplies the corrected data to an annular separator 5 for writing to the under control of the control circuit 6. The output of the ECC circuit 33 is also supplied to a detector 40 PSM. In the special reproduction mode, the PSM detector 40 detects the PSM information in the input sector from the read data of the disk 1 and supplies the detected PSM information to the control circuit 6. The control circuit 6 uses this PSM information to control the writing (or loading) of the I and P images in the annular spacer 5 in a special reproduction mode according to the information related to the number of bytes of deviation in the ip ipp descriptor to ensure that the length information in the flow data from an image I after the input sector to the second image P is written to the annular spacer 5. A focus control circuit (not shown) and a servo tracer circuit 8 controls the focus and the scanning of the reader 2, respectively, under the control of a system controller (not shown) according to a focus error signal and a tracking error signal obtained from the information n read by the reader 2.

Based on the sector address of each sector detected by the sector detector 4, the control circuit 6 designates, by means of a write indicator WP, a writing direction for writing the appropriate sector in the annular separator 5. Furthermore, in Based on a code inquiry signal obtained from a video code separator 10 (FIG. 8B), the control circuit 6 designates, by means of a reading indicator RP, a read address from which they are read. the data of the annular separator 5. The data read from the position designated by the read indicator RP is supplied to an optical switch 32. Since the encoded data recorded in the disk 1 comprise video data, audio data and subtitle data in multi-iplexed packets, the sub-iplexer 32 separates the supplied data into video data, audio data and subtitle data, and then supplies the respective data to a video cassette recorder. 20 (FIG. 8B), an audio decoder (not shown), and a sub-title decoder (not shown). As a result, the video data read from the annular separator 5 are stored in the video code separator 10 of the video decoder. The flow data from one I image to the second subsequent P-image contains packets other than the video packets, as shown in FIG. 2. In the special reproduction mode, any unnecessary data, ie, packets other than the video data is excluded by means of the iplexer demulting 32. The data stored in the video code separator 10 is supplied to an image header detector 34 where the image header, the image type information which means the image type I, P or B, and the temporal reference (TP.) which means the order of frames in the GOP is detected. The detected image type information is supplied to an image data selector 35, wherein only the I and P images are selected in the special reproduction mode; and the selected images are supplied to an inverted 11 VLC circuit. In the normal reproduction mode, the image data selector 35 is controlled to output all the image data without pre-selection. The data supplied to the inverse VLC circuit 11 is processed using inverse VLC; and then they are supplied to a decoder 12. The required code signals are returned to the video code separator 10 from the inverse VLC circuit to allow new data to be transferred from the video code separator 10. In addition, the inverse VLC circuit 11 sends a quantization step size to the signaling device 12 and sends a motion reader information to a motion compensator 15. E? 1 unquanti fi ed 12 discounted data recorded according to the size step of how much specified and sends the de-quantized data to an inverted DCT circuit 13. The inverse DCT circuit 13 processes the data quantized using reverse DCT and supplies the processed data to an adder 14. The adderadsr 14 adds the output of the inverse DCT circuit 13 and the output of the motion compensator 15 according to the type of image (I , P or B) and supplies the resulting motion compensated video data to a bank 16 of frame memory. Then, the read data from the frame memory bank 16 is rearranged in the original array of frames presented in FIG. 14A by means of the switch 16E. The real-time data is supplied to an analog numerical converter 17 (D / A) to be converted into an analog video signal which is then displayed on a display device 18. In response to a code inquiry signal that comes of the video code separator 10, the control circuit 6 supplies the data stored in the annular separator 5 to the video code separator 10. When the amount of data transferred from the video code separator 10 to the inverse VLC circuit 11 decreases, for example, as a result of the continuous processing of data from simple images having a small amount of data, the amount of data transferred from the annular separator 5 to the video code separator 10 also decreases. Accordingly, the amount of data stored in the annular separator 5 can be increased and the write indicator WP can potentially pass the read indicator RT to cause an overflow of the annular separator 5. To avoid such problem, the control circuit 6 calculates the current amount of data stored in the annular separator 5 based on the address positions of the write indicator WP and of the reading indicator RP. When the amount of calculated data exceeds a predetermined amount, a track skip decision circuit 7 determines that the annular spacer 5 may potentially be oversaturated and sends a track skip command to the servo track circuit 8. In response to the skip command of the track, the servo tracking circuit 8 causes the reader 2 to jump off the track according to the storage capacity of the annular separator 5 to avoid a supersaturation or a lack of flow of the annular separator 5. This allows the continuous reproduction of video with a quality image uniform, regardless of the continuity (or simplicity) of the images recorded on the disk 1. The data transfer rate from the annular separator 5 to the video code separator 10 is preset to be equal to or less than the data transfer rate of the ECC circuit 33 to the annular separator 5 to allow the video code separator 10 The separator is required to cancel a code scan for data transfer regardless of a track skip. In a normal reproduction mode it is considered, for example, that the image data I, P and B, 10, B-2, Bl, PO, BO, Bl, ... are recorded on disk 1 in the order presented in Figure 14B. In this example, a GOP is composed of 15 frames of images including a block of I image, 4 frames of images P and 10 frames of images B. The normal reproduction of the images is made by reading and unpacking the sequential ion of the images. data encoded in the order of registration presented in Figure 14B and displaying the decoded data in the order presented in Figure 14A. More specifically, at the time of decoding the image I, 10, the decoded output obtained from the inverse DCT circuit 13 is supplied directly to the frame memory bank 16. However, at the time of decoding the image B, B-2, the previously decoded P image (not illustrated) and the image I, 10, both used as references for the predictive coding of the image B, B-2, are supplied from the frame memory bank 16 to the motion compensator 15, and a predicted motion image is generated according to the information of the motion reader supplied from the inverse VLC circuit 11. The predicted image d? The generated movement is then supplied to the adder 14 which adds the predicted motion image to the output of the inverse DCT circuit 13, whereby the image B, B-2 is decoded and stored in the frame memory bank 16. The image B, Bl, is decoded in the same way as the image B, B-2, and overwrites the image B, B-2 overwritten in one of the frame memories of 16a-16c of the frame memory bank . In decaying the image P, PO, the image I, 10, is supplied from the frame memory bank 16 to the motion compensator 15 together with the information of the motion reader supplied from the inverse VLC circuit. The motion compensator supplies a predicted image of movement to the adder 14 that adds this predicted image of motion to the image P, PO, supplied from the inverse DCT circuit to decode the image P in this way, PO. The decoded image P, PO, decoded in this way is written on the above data (which can be either the image I or the image P) stored in the frame memory bank 16. Therefore, the images are sequentially decayed as mentioned above, but they are read from bank 16 of frame memory in their. The original order is displayed in the visualization device 18 in the order presented in Figure 14A. In a fast playback mode, the data recorded on disk 1 requires its decoding and display in reverse order. For example, if an image B, B07, is decoded from disk 1, on which video data is recorded in the order shown in figures 9A and 9B, (only video data is shown for simplicity), it is necessary decoding the images P, P08, and P05, before desiring the image B, B07, because these P images should be used as a reference to decode the desired image B, B07. However, it is noted that the decoded image P, P05, is required to decode the image P, POS, and the decoded image I, 102, is required to decay the image P, POS. Therefore, the decoding must start from the image I located at the beginning of the GOP. When the GOP decoding is finished, the operation requires jumping back to the previous GOP to continue the decoding process. However, if an inverse reproduction is made with such a decoding technique, an excessive temporal delay in the movement of the images will be introduced, which will result in the visualization of unnatural images. The present invention solves this problem by decoding only the I and P images as in the normal reproduction mode, but by performing reverse reproduction using only the same three memories 16a, 16b, and 16c of frames required for normal reproduction. Accordingly, the present invention mercifully deduct a total of three images, an I image appearing after an input sector and two subsequent P images, instead of decoding an entire sequence of I and P images.; consequently this allows a fast reverse reproduction with a minimum complexity of c. The present invention can perform fast reverse reproduction by detecting 40 PSM of a simplified circuit configuration without making use of a flow detector having a complicated configuration as described above. Since the PSM detector 40 detects information representing the number of bytes of deviation given in the ip ipp descriptor of a PSM, it is possible to write to an annular spacer 5 only the required range of the data flow from an I image appearing immediately after an entry sector until the two subsequent P images that appear after the I image. The operation performed by the data decoding apparatus of FIGS. 8A and 8B in a fast reverse reproduction (FR) mode will be described with reference to FIGS. 9A to 9C. Figures 9A and 9B show the order of video data recorded on the disc 1. The video data shown in these diagrams correspond to 4 GOPs, and in the reproduction mode FR, the control circuit 6 executes its control action of such that reader 2 reads the video data from disk 1 in the order indicated by the arrows below the video data. More specifically, the reader 2 successively reads the video data of the image I, 132, the image B, B30, the image P, P31, the image B, B35, the image B, B33, the image B, B34 and the image P, P38, in this order, then jumps to a immediately preceding GOP, and reads the video data from the image I, 122 to the second image P, P28. Subsequently, the reader 2 jumps to another previous GOP, and reads the video data from the image I, 112, to the second image P, P18. Then, the reader 2 jumps to a still more previous GOP, and then reads the video data from the image I, 102 to the second image P, POS. Subsequently, similarly to the foregoing, the reader 2 reads the video data of an I image positioned at the beginning of a GOP immediately before the second image P passed before this image I. Such a reading operation is possible because the information The above-mentioned input point is written into the input sector formed at the top of each GOP and the PSM detector 40 detects the ip ipp descriptor in the input sector and then supplies the detected ip ipp descriptor to the control circuit 6. Accordingly, the control circuit 6 controls the reader 2 to read, from the upper (or beginning) part of the input sector, the data corresponding to the number of bytes represented by the byte information up to the second image P given in the ip ipp descriptor, whereby the reader 2 can read the video data in the order indicated by the arrows in figures 9A to 9C. To have access to the upper part of the immediately preceding GOP, the distance information (data length) represented as a deviation direction in the PSD in the input sector - is used and means the distance to the previous input sector. The read video data that is located from the I image pasicianada at the beginning of the GOP to the second image P positioned after the image I are separated from the audio data and other data by means of the desmul t iplexor 32 and after it is written to the video code separator 10. The images B are eliminated by using the detection information obtained from the image header detector 34, and by decoding only the image data 3 and P which are written in bank 16 of frame memory. The video data written in this way is read from bank 16 of frame memory in the order of v is? The reverse image display presented in FIG. 9C and displayed in the display device 18. The reading / writing synchronization of the data for the frame memory bank 16 in the FP playback mode is deciphered. with reference to figure 10. Preferably, the frame memory bank 16 is equipped with 3 frame memories 16a, 16b and 16c, as shown in FIG. 8. The decoded image I is written, 132, in the upper part of the initial GOP (see FIG. 9b) in the frame memory 16a starts at the time tip tO and ends at the time point ti after a time duration of 1 frame. Subsequently, the writing of the image of P, P35, decoded with reference to the image I, 132, in the frame memory 16b starts at the point Pl, and ends at the time point t2 after a period of time of a frame. After the writing of the image P, P32, decoded with reference to the image P, P35, in the frame memory 16, it starts at the time point t2, and ends at e3 time point t3 after a period of time of a picture. The reading of the image P, P38, from the frame memory 16c can start at an intermediate point between the two time points t2 and t3, provided that a field of the image P, P38, has already been written in memory of frames 16c at this time point of time start. Accordingly, it is possible to read and write simultaneously in the same frame memory by delaying read synchronization compared to write synchronization by a field. The reading of the image of P, P38, from the frame memory 16c ends at an intermediate point between the two time points t3 and t4, and the writing of the decoded image I, 122, of the preceding GOP in the frame memory 16c starts at time point t3. This writing ends at time point t4 after a period of one frame. Different image data can be described in the frame memory 16c while simultaneously reading the previously written image data according to the aforementioned because the write synchronization preferably has a delay of a field in relation to the read synchronization. Then, the decoded image data as shown in Figures 9A to 9B are written into the memories of S '? tables 16a, 16b and 16c, in the order of 132, P35, P38, 122, P25, P28, 112, P15, P18, 102, POS etc. While reading the data from? image from the frame memories 16a, 16b and 16c in the order from the oldest image number (largest) to the most recent image number (smallest) as P38, P35, 132, P28, P25, 122 , P18, P15, 112 ... et. Accordingly, fast reverse reproduction is performed and the figures are displayed in the order presented in Figure 9C. For example, 3 images per TOP can be reproduced inversely when 3 frame memories are used. In the reverse reproduction mode, identification numbers assigned to the images are detected, and it is seen in the images from the bank 16 of frame memory in the order from the oldest number to the most recent number (from the highest to the smallest ). Temporary references (TR), which means the numbers that indicate the display order of the images, are placed back in the respective upper parts of the GOPs, and the values of such temporary references are within a range of 0 to 1023 With reference to Figures 11A 11C, an explanation for fast forward reproduction (FF) performed in the data decoding apparatus of Figures 8A and 8B is provided. Figures HA and 11B show the video data of 4 GOPs in the order in which they are recorded on disk i, where the arrows under the video data indicate the order of reading of the video data in the reproduction mode. FF. In the FF replay mode, the PSM detector 40 detects u.rs ip ip descriptor in the input sector written at the top of each GOP and supplies the detected ip ipp descriptor to the control circuit 6, as in the mode of reproduction FR mentioned above. The control circuit 6 controls the reader to read, from the top of the input sector, the data corresponding to the number of bytes represented by the byte information to the second image P in the ip ipp descriptor, so that the video data are read in the order indicated by the arrows in figure 11 C. The images B, whose identities are present in the respective headings of the images, are removed from the video data read in such a way that they are only decoded the I and P images. The decoded images I and P are read from the bank 16 of frame memories in the descrambling order and are displayed in the display device 18 in the order of 102, POS, P08. 112, P15, P18, 122, P25, P28, 132, P35, P38, as shown in Figure 11C.

Even if three frames are preferably incorporated in the frame 16 of frame memory, the number of frame memories is not limited to three, any desired number can be chosen. The fast playback hp ia front (FF) will be done with a number of I and P images equal to the frame memories. The FR reproduction made with only two frame memories 16a and 16b in the frame memory bank 16 will be described with reference to FIGS. 3A to 12C and FIG. 13. FIGS. 32A and 12B show the burn in which the frames are recorded. video data on the disk 1. The video data in FIGS. 32A and 12B correspond to four GOPs, and in the FP-play mode, the control circuit 6 controls the reader to read the video data from the disk 1 ep the order indicated by the arrows below the video data. More specifically, the reader 2 successively reads the video data of the image I, 132, image B, B30, image B, B31 and image P, P35, in this burst, then jumps to an immediately previous GOP, and reads the video data of the image I, 122, up to the first image P, P25. Subsequently, the reader 2 jumps to another previous GOP, and reads the image video data I, 112 to the first image P, P15. Then, reader 2 jumps to a preceding GOP, and then reads the image video data I, 102 to the first image P, POS. Then, similarly to the foregoing, the reader 2 reads the video data from an image I positioned at the beginning of the immediately preceding GOP until the first image placed after this image I. Such operation is performed because the detector 40 PSM detects the information of bytes to the first image P in an ip ipp descriptor in the input sector written at the top of each GOP and then supplies the detected information to the control circuit 6. More specifically, the control circuit 6 controls the reader 2 to read, from the p >Higher art of the input sector, the data correspond to the number of bytes represented by the byte information up to the first image P given in the ip ipp descriptor, so that the reader 2 can read the video data in the order indicated by the arrows in fig. 9A to 9B. The read video data that are located from the image I positioned at the beginning of the GOP to the first image P positioned after this image I are separated from the audio data and other data by means of the demultiplex 32 and are written to the separator Subsequently, the P images, whose identities are present in the respective image headers, are deleted and only the image data I and P are decoded and written to the frame memory bank 16. The video data written in this way is read from the frame memory 16 in the order of image display shown in FIG. 12C and is displayed on the display device 18. FIG. 13 is a data read / scan synchronization diagram for bank 16 of frame memory of a capacity of two blocks. The writing of the decoded I image, 132, at the beginning of the most recent GOP in FIG. 12B in the frame memory 16A starts at the time point PO, and ends at the time point Pl, after a period of one frame. Subsequently, the writing of the image P, P35, decoded with reference to the image I, 132, in the frame memory 16B starts at the point ti, and ends at the time point t2 after a period of one frame. The reading of the image P, P35, starting from the frame memory 16B starts at an intermediate point between the two temporary points ti and t2 provided that a field of the image P, P35, has already been written in the memory of frames 16b at this temparal starting point of reading. Accordingly, it is possible to read and write simultaneously in the same frame memory 16b by delaying read synchronization in relation to write synchronization by a field. Accordingly, different image data may be written to the frame memory 16b while simultaneously viewing previously written image data. The reading of the image P, P35 from the frame memory 16b ends at an intermediate point between the two time points t2 and t3, and the writing of the de-graded image I, 122, of the preceding GOP in the frame memory 16b starts at time point t2. This writing ends at time point t3 after a period of one frame. The image I, 132, is seen from the frame memory 16a at an intermediate point between the two time points t2 and t3, and after about a delay of a field from the initial point of reading, the writing of the P-decoded image P25 in the frame memory 16a starts. A frame of the image I, 132 is read entirely at an intermediate point between the time points t3 and t4, and subsequently a frame of the image P, P25 is completely read from the frame memory 16a. In addition, a frame of the image I, 13.2, in the previous GOP is read in the frame memory 16a between the time points t4 and tS. The desired image data as shown in FIGS. 12A to 12B are written in the frame memories 16a and 16b in the order 132, P3S, 122, P25, 112, P15, 102, POS ...; and they are read in the order from the oldest image number (largest number) to the most recent image number (smaller number) such as P35, 132, P25, 122, P15, 112, POS, 102 ... As described above , an I image and two or a P image by GOP are presented in the special playback mode. It will be appreciated that the present invention can be modified to decode and display an image I by GOP and to delete both P or B images. In this case, the information to detect the number of bytes up to the end of image I is register in a PSD (Program Flow Directory). In more specific form, in the program flow directory defined according to the MPEG system (ISO 13818-1), the information relating to the image I immediately after the PSD is recorded as the reference access unit, and three position deviation values are added. PES header, reference deviation and bytes to read together to determine the data length (total number of bytes) from the first byte of the PSD to the end of image I. When the storage capacity of the frame memories exceeds In three frames, more than three frames can be decoded by GOP and can be played in special playback mode. In such a case, the information representing the data length is written in PSM in such a way that three or more P images appear after the I image has been accessed.

Even though the previous examples show a jump to an adjacent GOP in the special replay mode, the skip can be made to a distant GOP when performing a special playback. In the present invention where the reader 2 jumps in the special playback mode, the video data in the video stream is at different speeds since the compression degrees thereof differ according to the type of image (I, P or B) ) or the attributes of these images (flat or complex). As a result, the search time is not fixed and some difficulties may arise in performing the FF / FR speed reproduction. To avoid such difficulties, the search time or visualization interval is measured by means of the system controller and the distance to the next search is changed according to the time edited, so speed control is achieved through of the control of the retraction imentación. For example, if a longer search time has been taken, reader 2 jumps to a slightly distant position in a GOP unit to gain the necessary distance in this way. Although this invention has been described in relation to an optical recording medium, it should be understood that the recording method and apparatus called reproduction of the present invention can be employed with other recording means or for transmitting the compressed video data. Since special reproduction for example, reverse reproduction can be carried out by the present invention with a circuit of lower cost and complexity, such components of the apparatus that include substrates and special circuits can be reduced in terms of their measurements, which consequently decreases the consumption of energy, thus minimizing the generation of heat and minimizing the structure necessary for heat radiation. In the special playback mode, a single I image can be recovered, one I image and one P image can be recovered, or one I image and two P images can be recovered. These settings can be selectively interchanged in such a way that the velocity Special playback can be controlled by changing the number of images to be loaded into the frame memory and displayed. By using one or two P images in addition to an I image, scenes are obtained that can be displayed smoothly to provide a satisfactory visual presentation. While the present invention has been described and presented particularly with reference to the preferred embodiments, it will be readily appreciated that various changes can be made without departing either from the spirit or from the scope of the present invention. It is intended that the appended claims be construed to include the aforementioned embodiments, the numerous alternatives that have been described and all their equivalents.

Claims (37)

1. CLAIMS: 1. A method for recording image data in a recording medium, comprising the steps of: receiving said image data; encoding said image data using intraframe coding and predictive coding to provide a data stream containing at least one intraframe encoded image data (I image) and at least two predictive encoded image data (P images) in a predetermined order , where said at least one image I precedes said at least 2 P images; generating position information representing the positions of said at least one I image and said at least two P images in relation to said at least one I image in said data stream; registering said at least one I image, said when you give images P and said position information on said recording medium. The method according to claim 1, wherein the step of generating position information includes the generation of position information representing the data length of said at least one I image until the end of a first one of said at least two P images, and from said at least one I image until the end of a second of said at least two P-3 images. The method according to claim 2, wherein the register step registers said when less than one image I and said at least 2 P images in a sector, and wherein the step of generating position information further includes generating information of position representing the length of data from the beginning of said sector including said at least one I image until the end of said first image P, and starting from the beginning of said sector until the end of said second image P. 4. The method according to claim 3, wherein the registration step includes recording said position information in said sector, 5. The method according to claim 4, wherein the coding step includes bidirectional predictive coding of the image data. to provide a plurality of bidirectionally predictive encoded images (image B) in said data stream. 6. The method according to claim 4, wherein each of said at least one image I and said at least two images P constitute a frame. 7. An apparatus for recording image data in a recording medium, comprising: means for receiving such image data; means for encoding said image data using intraframe coding and predictive coding to provide a data stream containing at least one intraframe encoded image data (I image) and at least two predictive encoded image data (images P) in a predetermined order, where said at least one I image precedes said at least two images means for generating position information representing the positions of said at least one I image and said at least two P images in relation to said at least one I image in said data stream; and means for registering said at least one I image, said at least two P images and said position information in said recording means. The apparatus of claim 7, wherein said means for generating can be operated to generate position information representing the data length from said when an I image drops to the end of a first one of said at least two P images, and from said at least one I image until the end of a second one of said at least two images P. 9. The apparatus of claim 8, said means for registering can be operated to record said at least one I image and said when minus two P images in a sector; and wherein said means for generating can be operated to generate position information further representing data length from the beginning of said sector including said at least one I image until the end of said first image P, and from the beginning of said sector until the end of said second image P. 10. The apparatus of claim 9, wherein said means for registering can be operated to register said position information in said sector. The apparatus of claim 10, wherein said means for coding can be operated to encode the image data using the bidirectionally predictive coding of the image data to provide a plurality of images encoded in predictive predictive bidirectional mode (image B) in said data flow. The apparatus of claim 10, wherein each of said at least one image I and said at least two images P constitute a frame. 13. A method for reproducing image data from a recording medium, said image data is recorded as a data stream representing groups of images (GOPs), each GOP is composed of at least one intraframe encoded image data (image I) and at least two image data encoded in predictive form (P &g images);, wherein said at least one I image precedes said when you mine two P images, and said method comprises the steps of: reproducing said data stream; detecting position information included in said data flow and representing said positions at least; an I image and said at least two P images in relation to said at least one S image in said data stream; detecting from said data flow said at least one I image and said when you give P images in response to said position information and deriving coded image data therefrom; decoding said encoded image data to provide decoded image data; and display said decoded image data. The method according to claim 13, wherein the step of detecting includes detecting position information information representing data length from said at least one I image to the end of a first one when said two P images, and from said at least one I image until the end of a second one of said at least two images P. 15. The method according to claim 14, wherein the step of reproducing reproduces said data stream, from of a sector; and wherein the detecting step further includes detecting position information representing data length from the beginning of said sector including said at least one I image until the end of said first image P, and from the beginning of said sector up to the end of said second image P. 16. The method according to claim 15, wherein said position information is recorded in said sector. 17. Apparatus for reproducing image data from a recording medium, said image data is recorded as a data stream representing groups of images (GOPs) with each GOP composed of at least one intraframe encoded image data (image I) and at least two image data encoded in predictive form (P images), where said at least one I image precedes said at least two P images, said apparatus comprises: an image data, which captures: means for reproducing said medium data stream for detecting position information included in said data flow and representing position of said at least two P images in relation to said at least one I image in said data stream; means for detecting from said data stream said at least one I image and said at least two P images in response to said position information and deriving encoded image data therefrom; means for decoding said encoded image data to provide decoded image data; and means for displaying said desired image data. The apparatus of claim 17, wherein said means for detecting can be operated to detect position information representing data length from said at least one I image to the end I of a first one of said at least two P images, and thereafter when you indicate an image I until the end of a second one of said at least two images P. 19. The apparatus of claim 18, wherein said reproducing means can be operated to reproduce said data stream from a sector; and wherein said means for detecting said position information can be further operated to detect position information representing data length from the beginning of said sector including said at least one I image until the end of said first image P, and from the beginning of said sector until the end of said second image P. 20. The apparatus of claim 19, wherein said position information is recorded in said sector. The apparatus of claim 20, wherein said image data encoded in the stream of reproduced data includes various image data encoded in bidirectionally predictive form (images B), 22. The apparatus of claim 20, wherein each of said at least one I image and said at least two images P cans itutye? n picture. 23. A means of registration for use in combination with the • processor-controlled apparatus and having an image data encoded or registered in thesaid encoded image data comprises at least one intraframe encoded image (image I), and at least two images encoded in predictive form (P images), as well as position information representing respective positions of said at least one I image and said at least two P images in relation to said at least one I image, thus enabling the processor controlled apparatus to reproduce the image data encoded in a special reproduction mode from the recording media by using the position information for selecting the image encoded image data P. 24. The recording medium of claim 23, wherein said position information represents data length from said at least one I image to the end of a first of said at least two P images, and from said at least one I image to the end of a second of said at least two images P. 25. The recording medium of claim 24, wherein said encoded image data is recorded in a sector; and wherein said positional information additionally represents data length from the beginning of said sector including said at least one image I. until the end of said first image P, and from the beginning of said sector to the end of said second image P. 26. The registration means of claim 25, wherein said position information is recorded in said sector. 27. The recording medium of claim 26, wherein said encoded image data includes several pictures in a predictive manner (images B) 28. The recording medium of claim 26, wherein each of said when less an image I and said at least two P images constitute a picture. 29. Apparatus for recording and reproducing image data in a recording medium, comprising: means for receiving said image data; means for encoding said image data using intraframe coding and / or predictive coding to provide a data stream containing at least one intraframe encoded image data (image I) and at least two image data encoded in predictive form (images) P) in a predetermined order, wherein said at least one I image precedes said at least two P images; means for generating position information representing positions of said at least one I image and said at least two P images in relation to said at least one I ep image said data stream; means for registering said at least one I image, said when less than two P images, and the position information in said recording means; means for reproducing said data flow from said recording means; means for detecting said position information from said reproduced data stream; means for detecting said said data flow when an I image is undershot and said when two P images in response to said detected position information and deriving coded image data therefrom; means for decoding said encoded image data to provide decoded image data; and means for displaying said desired image data. 30. The apparatus of claim 29, wherein said means for generating position information can be operated to generate position information representing the data length from said at least one I image to the end. from a first of said at least two images P, and from said at least one image I to the end of a second of said at least two images P. 31. The apparatus of claim 30, wherein said means for recording can be operated to register said at least one I image and said at least two P images in a sector; and wherein said means for generating can be operated to generate position information that additionally represents data length from the beginning of said sector including said at least one I image until the end of said first P image, and from the beginning of said sector up to the end of said second image P. 32. The apparatus of claim 31, wherein said means for recording can be operated to register said position information in said sector. The apparatus of claim 32, wherein said means for coding can be operated to encode the image data using the bidirectionally predictive coding of the image data to provide several images encoded in predictive binary form (images B) in said data flow. 34. The apparatus of claim 32, wherein each of said at least one I image and said when less than P images constitutes a frame. 35. The apparatus of claim 29, wherein said means for detecting can be operated to detect position information representing data length from said at least one I image until the end of a first of said at least two P images, and starting from said at least one I image to the end of a second one of said at least two images P. 36. The apparatus of claim 35, wherein said reproducing means may be operated to reproduce said data stream from a sector; and wherein said means for detecting said position information is further operated to detect position information representing data length from the beginning of said sector including said at least one frame I up to the end of said first image P, and starting from the beginning of said sector until the end of said second image P 37. The apparatus of claim 36, wherein said poεition information is recorded in said sector.