JP2003274412A - Image encoding method and image decoding method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To enhance the compression rate of image encoding. <P>SOLUTION: A Memalloc 1 compares a reference image number Idx with a short time storage area size Stsize and with a long time storage area size Ltsize and a long time storage indication flag LTflg informs of whether the reference image number Idx indicates an image in a short time storage area or an image in a long time storage area. When the Idx exceeds an Idx value within a range decided by the Stsize and the Ltsize, an extended motion compensation indication flag AdvPred informs of a result of applying an arithmetic operation to a plurality of reference images for a reference image. A reference image object picture type StrPTYPE is a picture type when the image stored in a frame memory Mem 3 is encoded, and when the StrPTYPE for encoding a block prediction type MBtype is a picture type impossible for reference, no correction reference image number RIdx is assigned to the StrPTYPE. <P>COPYRIGHT: (C)2003,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image coding method for efficiently compressing a moving image signal by utilizing the correlation between screens, an image decoding method for correctly decoding the same, and a software implementation thereof. It is a recording medium in which a program for is recorded.

[0002]

2. Description of the Related Art In recent years, in the age of multimedia in which voices, images, and other pixel values are handled in an integrated manner, conventional information media such as newspapers, magazines, televisions, radios, and telephones are transmitted to people. Means have become the subject of multimedia. In general, multimedia means not only characters but also figures, voices, especially images, etc. are associated with each other at the same time. However, in order to target the above-mentioned conventional information media as multimedia, the information is converted into a digital format. It becomes an indispensable condition.

However, when the information amount of each of the above information media is estimated as a digital information amount, the amount of information per character is 1 to 2 bytes in the case of characters, whereas that of voice is 64 kbits per second. (Telephone quality), and for moving images, an information amount of 100 Mbits (current television reception quality) or more is required per second, and it is not realistic to handle the enormous amount of information in the digital form as it is in the above information media. For example, videophone is 64kbps ~ 1.5M
Integrated service digital network (ISDN: bps)
It has already been put to practical use by the Integreted Services Digital Network), but it is impossible to send the video of the TV / camera as it is with ISDN.

Therefore, what is needed is an information compression technique. For example, in the case of a videophone, it has been internationally standardized by ITU (International Telecommunication Union Telecommunication Standardization Sector).
H.261 and H.263 standard video compression technology is used. In addition, according to the MPEG-1 standard information compression technology,
It is also possible to put image information along with audio information on a CD (compact disc).

Here, MPEG (Moving Picture Experts G
roup) is an international standard for digital compression of moving picture signals, and MPEG-1 is a standard for compressing moving picture signals up to 1.5 Mbps, that is, about 1 / 100th of the information of television signals.
In addition, the transmission rate for the MPEG-1 standard is mainly about 1.
Since it is limited to 5 Mbps, the moving picture signal is compressed to 2 to 15 Mbps in MPEG-2 standardized to meet the demand for higher image quality.

Furthermore, at present, a working group that has been promoting standardization with MPEG-1 and MPEG-2 (ISO / IEC JTC1 / SC29 / WG11)
Has standardized MPEG-4, which has a higher compression rate. M
Initially, PEG-4 not only enables highly efficient encoding at a low bit rate, but also introduces powerful error resilience technology that can reduce subjective image quality degradation even if a transmission path error occurs. . In addition, ISO / IEC and ITU are jointly working on the standardization of JVT (Joint Video Team) as a next-generation screen encoding method, and at the present time, Joint Model 1 (JM
The one called 1) is the latest.

In JVT, unlike the conventional moving picture coding,
An arbitrary image (frame) can be selected as a reference image from a plurality of images (frames) as the forward reference image. Figure
10 is an explanatory diagram of image encoding using a short-time storage memory. The short-term storage memory stores several images decoded immediately before, and refers to so-called MPEG-1 or MPEG-2 P pictures (forward predictive coded pictures) and B pictures (bidirectional predictive coded pictures). Corresponds to the image.

FIG. 10A shows an example of P picture prediction.
In P picture, it is possible to refer to only I picture or P picture. Frame 3 refers to Frame 0, Frame 6 refers to Frame 0 or Frame 3, Frame 9
Encodes with reference to Frame 0, Frame 3 or Frame 6
Can be decrypted.

FIG. 10B shows an example of B picture prediction.
With B pictures, I pictures and P pictures can be referred to as forward pictures. The rear image can refer to one screen that is temporally closest to the I picture or P picture. Frame 1 and Frame 2 forward refer to Frame 0, Frame 4 and Frame 5 refer to Frame 0 or Frame 3, Frame 7 and Frame 8 refer to Frame 0, Frame 3 or
Encoding / decoding can be performed by referring to Frame 6.

Note that in JVT B picture prediction, I
In addition to pictures and P pictures, it has been introduced to refer to B pictures as forward pictures.

FIG. 11 is an explanatory diagram of image coding using a long-term storage memory. The short-term storage memory stores several images immediately (encoded and decoded) temporally immediately before, while the long-term storage memory explicitly stores and stores the image regardless of the passage of time. Is to be destroyed. For example, when switching between multiple cameras with a surveillance camera, if the final decoded image for each camera is stored in a long-term storage memory, the same camera can be used to capture the last image when the same camera is used again. If the difference value from the image stored in the time storage memory is encoded, the correlation between the images is high and the compression rate can be improved.

As a reference image of the target image CurFrame in FIG. 11, LTFrame 0, LTFrame 1, LTFrame 2, LTFrame 3 which are screens older than the short-term storage memory shown in FIG. 10 are displayed.
You can see that you are referring to. This LTFrame 0, LTFrame
1, LTFrame 2 and LTFrame 3 are stored in the long-term storage memory.

FIG. 12 is an explanatory diagram of a conventional decoded image management method. FIG. 12A shows a state in which the short-term storage memory and the long-term storage memory are secured in the frame memory that is the reference image memory. The total number of usable images (frames) for both short-term memory and long-term memory is JV.
The value is fixed in the T plan, and is set to 13 in the example of FIG. Therefore, in JVT, the long-term storage area size LT
By encoding size, the short-term storage area size STsize
= Total number of usable images-Short-time storage area size STsize is obtained as LTsize. A reference image number Idx is assigned to each area in which the reference image is stored, and the reference image is identified by this reference image number Idx. Further, although the short-term storage memory and the long-term storage memory have an area (unused) in which an image is not yet stored, the reference image number Idx is assigned without making a distinction from the area (used) in which the image is stored.

When encoding / decoding by referring to an image, the image referred to as the forward image is represented by the reference image number Idx.

FIG. 12 (b) shows a method proposed in the JVT standardization for identifying the extended predictive coding by using the reference image number Idx. If a number (indicated by AdvPred) that exceeds the reference image number Idx in the range determined by the JVT proposal is used as the reference image number Idx indicating the reference image, refer to the interpolated average value of the previous two screens as the reference image. Yes (Type1 Average) Use (Type2 Average) to refer to the extrapolated predicted values of the previous two screens. It has been reported that the coding efficiency is improved by referring to the interpolated average value or the extrapolated predicted value when the luminance linearly changes with time in a moving image.

FIG. 9 is a block diagram of an image coding apparatus using a conventional image coding method. The operation will be described below. The picture type PTYPE indicates whether the picture is a P picture that refers to only the forward image or a B picture that can also refer to the backward image in motion detection, and limits the images that can be referenced by the motion detector ME. The motion detector ME compares the image signal Vin with the motion detection reference image MEpel, and outputs the reference image number Idx indicating the reference image having the smallest difference value between them and the motion vector MV that is the amount of movement of the pixel. In addition, whether the image referred to at that time is a front image, a rear image, or an average value of both images is a block prediction type MBty.
Notify with pe.

The memory access controller MemAlloc1 compares the reference image number Idx with the long-term storage area size LTsize,
The long-time storage instruction flag LTfl indicates whether the reference image number Idx indicates an image in the short-time storage area or an image in the long-time storage area.
Notify with g. If the reference image number Idx exceeds a number that exceeds the reference image number Idx in the range determined by the JVT proposal, the extension motion compensation instruction flag AdvPred should be used as the reference image that is the result of calculation of multiple reference images. Notify with.

The frame memory Mem1 has a reference image number Idx,
Long-term memory instruction flag LTflg, extended motion compensation instruction flag A
The pixels in the area indicated by dvPred and the motion vector MV are output as the motion compensation reference image Refpel, and the motion compensator MC outputs the extended motion compensation instruction flag AdvPred (calculation using multiple images) and the block prediction type MBtype (for both images). The calculation indicated by (average) is performed and the motion compensation image MCpel is output.

The subtractor Sub receives the image signal Vin and the motion compensation image MC
The residual image Dif, which is the difference value of pel, is calculated, and the orthogonal transformer T
The frequency component Coef converted into the frequency domain by is output. The quantizer Q quantizes the frequency component Coef and outputs a quantized value DecQ.

The inverse quantizer IQ dequantizes the quantized value DecQ and outputs the decoded frequency component DecCoef, and the inverse orthogonal transformer IT outputs the decoded frequency component DecCoef to the decoded residual image which is the difference value of pixels from the frequency domain. Output DecDif. The adder Add1 adds the decoded residual image DecDif and the motion compensated image MCpel to obtain the decoded image signal Decpel, and stores it in the frame memory Mem1.
On the other hand, the variable-length encoder VLC has a motion vector MV, a reference image number Idx, a block prediction type MBtype, a long-term storage area size LTsize, a picture type PTYPE, and a quantization value D.
ecQ is encoded and output as encoded signal Str1.

As described above, an explanatory diagram of image encoding using the short-term storage memory of FIG. 10, an explanatory diagram of image encoding using the long-term storage memory of FIG. 11, and a conventional decoded image of FIG. An encoding device using the reference described in the explanatory diagram of the management method can be realized.

FIG. 13 is a block diagram of a memory access controller MemAlloc1 for implementing the conventional image coding method. The operation will be described below. Long-term storage area size
LTsize is a long-term storage buffer size calculator LTbuf
The long-term storage area size LTsize1 having the same value as the long-term storage area size LTsize is output until the long-term storage area size LTsize is updated.

The corresponding buffer determiner BankSel compares the reference image number Idx with the long-term storage area size LTsize1 and determines whether the reference image number Idx is within the long-term storage memory range indicated by the long-term storage area size LTsize1. Notify by the time memory instruction flag LTflg. In addition, the short-term storage area size STsize and long-term storage area size LT determined by the JVT plan
When the reference image number Idx exceeds the sum of sizes, the extended motion compensation instruction flag AdvPred indicates that the calculation results of a plurality of images are referred to.

FIG. 14 is a block diagram of an image decoding apparatus using the conventional image decoding method, and the code encoded by the block diagram of the image encoding apparatus using the conventional image encoding method of FIG. The decoded signal Str1 is decoded. In FIG.
Devices that operate in the same manner as in the block diagram of the image encoding apparatus using the conventional image encoding method are attached with the same numbers, and description thereof is omitted.

The variable length decoder VLD decodes the encoded signal Str1 and outputs a picture type PTYPE, a motion vector MV, a reference image number Idx, a long-term storage area size LTsize, a block prediction type MBtype, and a quantized value DecQ. .

The operations of the memory access controller MemAlloc1, the inverse quantizer IQ, and the inverse orthogonal transformer IT are the same as the block diagram of the image encoding apparatus using the conventional image encoding method of FIG. The frame memory Mem2 has a reference image number Idx, a long-time storage instruction flag LTflg, an extended motion compensation instruction flag AdvPred,
The pixels in the area indicated by the motion vector MV are output as the motion compensation reference image Refpel, and the motion compensator MC outputs the extended motion compensation instruction flag AdvPred (calculation using multiple images) and the block prediction type MBtype (average of both images). The motion compensation image MCpel is output by performing the operation indicated by.

The adder Add2 adds the decoded residual image DecDif and the motion-compensated image MCpel to output a decoded image signal Vout, and stores it in the frame memory Mem2 for reference in decoding the subsequent image. With the above configuration, the encoded signal Str1 encoded in the block diagram of the image encoding device using the conventional image encoding method of FIG. 9 can be correctly decoded.

[0028]

In such a conventional image coding method and image decoding method, the sum of the short-term memory and the long-term memory is constant and cannot be changed. Even if only a small-sized short-time memory and a long-time memory are used originally, a large reference image number Idx is necessary to indicate that the calculation result of multiple images is referred, and the coding efficiency is high. Does not improve. Furthermore, since the reference image number Idx is also assigned to the unreferenceable image in the frame memory, the encoding efficiency does not improve when the reference image number Idx is encoded.

Therefore, the present invention provides an image coding method and an image decoding method in which the compression rate is higher than that of the conventional image coding method and image decoding method by coding the reference image number Idx more efficiently. The purpose is to provide.

[0030]

In order to solve this problem, a first invention is an image coding method for coding by referring to a predetermined number of images selected from a plurality of reference images,
A buffer size indicating a buffer area for storing a reference image is determined, the determined buffer size is encoded, and if the index indicating the image to be referred indicates an image stored in the buffer area, refer to the indicated image, When the index indicating the image to be referred indicates outside the buffer area, the image encoding method is for encoding by referring to the result of the operation on the plurality of images stored in the buffer area.

A second invention is an image decoding method for decoding by referring to a predetermined number of images selected from a plurality of reference images, wherein a buffer size indicating a buffer area for storing the reference image is decoded, When the index indicating the image to be referred indicates the image stored in the buffer area, the image is referred to, and when the index indicating the image to be referred indicates outside the buffer area, the image is stored in the buffer area. It is an image decoding method for decoding by referring to the result of calculation on a plurality of images.

[0032]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described below with reference to FIGS. (Embodiment 1) FIG. 1 is a block diagram of an image coding apparatus using the image coding method of the present invention, and is the same as the block diagram of the image coding apparatus using the conventional image coding method of FIG. Devices that operate are given the same symbols, and explanations are omitted.

The operation of the block diagram of the image coding apparatus using the image coding method of the present invention shown in FIG. 1 will be described below. The motion detector ME compares the image signal Vin with the motion detection reference image MEpel, and the reference image number Idx indicating the reference image with the smallest difference value between the two and the motion vector M that is the amount of movement of the pixel.
Output V. Further, the block prediction type MBtype is used to notify whether the image referred to at that time is a front image, a rear image, or an average value of both images.

The memory access controller MemAlloc1 compares the reference image number Idx with the short-term storage area size STsize and the long-term storage area size LTsize, and the reference image number Idx indicates an image in the short-term storage area or the long-term storage area. Is displayed by the long-time storage instruction flag LTflg. If the reference image number Idx exceeds a number that exceeds the reference image number Idx in the range determined by the short-term storage area size STsize and the long-term storage area size LTsize, refer to the result of performing the operation on multiple reference images. The extended motion compensation instruction flag AdvPred is used to notify that the image is an image. Further, the corrected reference image number RIdx assigned only to the images that can be referred to is output to the variable length encoder VLC. The reference image candidate picture type StrPTYPE is a picture type when the image stored in the frame memory Mem3 is encoded, and the reference image candidate picture type StrPTYPE cannot be referred to when the block prediction type MBtype is encoded. In the case of a picture type, the modified reference image number RIdx for the reference image candidate picture type StrPTYPE is not assigned.

The frame memory Mem3 has a reference image number Idx,
Long-term memory instruction flag LTflg, extended motion compensation instruction flag A
The pixels in the area indicated by dvPred and the motion vector MV are output as the motion compensation reference image Refpel, and the motion compensator MC outputs the extended motion compensation instruction flag AdvPred (calculation using multiple images) and the block prediction type MBtype (for both images). The calculation indicated by (average) is performed and the motion compensation image MCpel is output.

On the other hand, the variable-length encoder VLC1 has a motion vector MV, a modified reference image number RIdx, and a block prediction type MBty.
pe, the short-term storage area size STsize, the long-term storage area size LTsize, the picture type PTYPE, and the quantized value DecQ are encoded and output as the encoded signal Str2.

In the memory access controller MemAlloc3, the short-term storage area size STsize and the long-term storage area size
Since the extended motion compensation instruction flag AdvPred is determined by whether the reference image number Idx exceeds the range determined by both LTsize, it is not within the upper limit determined by the JVT proposal but within the range of the storage area actually used. Since a small reference image number Idx can be used in, the number of bits required for encoding the reference image number Idx, that is, the modified reference image number RIdx can be reduced. When the reference image candidate picture type StrPTYPE is a picture type that cannot be referenced when encoding with the block prediction type MBtype, the modified reference image number RIdx for the reference image candidate picture type StrPTYPE is not assigned, so the modified reference image number The number of bits required for RIdx encoding can be reduced.
The corrected reference image number RIdx is assigned only to the area where the image is actually stored in the storage area, and the corrected reference image number RIdx for the area where the image is not stored (unused)
Is not assigned, the number of bits required for encoding the modified reference image number RIdx can be reduced.

FIG. 2 is an explanatory diagram of a decoded image management method when B pictures are not stored in the frame memory. The reference image number Idx is assigned to all areas in which images can be stored, but the modified reference image number RIdx does not assign numbers to areas (unused) where images are not actually stored. Therefore, the number of bits can be reduced by encoding the modified reference image number RIdx rather than encoding the reference image number Idx.

FIG. 3 is an explanatory diagram of a P-picture encoding decoded image management method when B-pictures are stored in the frame memory. Since the B picture can be referred to in the encoding of the B picture in the JBT, the B picture is also stored in the storage area. However, since the reference of the B picture is prohibited in the P picture encoding, the modified reference image number RIdx is not assigned to the B picture in the storage area during the P picture encoding. Therefore, considering that the number is not assigned to the area in which the image is not actually stored, the corrected reference image number RIdx is assigned as shown in FIG.

FIG. 4 is an explanatory diagram of a decoded picture management method for B picture coding when B pictures are stored in the frame memory. Only the front image can be selected as the reference frame,
Rear images cannot be selected. Therefore, the corrected reference image number RIdx is not assigned to the target image whose display time as the rear image. Therefore, considering that the number is not assigned to the area where the image is not actually stored, the modified reference image number
RIdx allocation is shown in Figure 4.

(Second Embodiment) FIG. 5 shows a memory access controller MemAlloc3 for implementing the image coding method of the present invention.
It is a block diagram of. In the block diagram of the memory access controller MemAlloc3 for realizing the image coding method of the present invention in FIG. 5, the same operation as the block diagram of the memory access controller MemAlloc1 for realizing the conventional image coding method in FIG. The same number is assigned to the device that does the same, and the description is omitted.

The short-time storage area size STsize and the long-time storage area size LTsize are input to the short-time storage buffer size calculation holder STbuf and the long-time storage buffer size calculation holder LTbuf, and the short-time storage area size STsize and Until the long-term storage area size LTsize is updated, the short-term storage area size STsize and the short-term storage area size STsize with the same value as the long-term storage area size LTsize
Output 1 and long-term storage area size LTsize1.

Unused buffer INDEX deletion corrector ModIdx determines whether or not the image in the area corresponding to the reference image number Idx is stored, reference image candidate picture type StrPTYPE
The corrected reference image number RIdx is not assigned to the area where the image is not stored. Also, block prediction type
MBtype is compared with the reference image candidate picture type StrPTYPE to determine whether the image indicated by the reference image candidate picture type StrPTYPE can be referenced with the block prediction type MBtype. If it is not possible, the modified reference image number RIdx is not assigned. .

The corresponding buffer determiner BankSel1 compares the reference image number Idx with the short-term storage area size STsize and the long-term storage area size LTsize1, and the reference image number Idx is indicated by the short-term storage area size STsize. Long-term storage area size LT Long-term storage instruction flag LT indicating whether it is within the range of long-term storage memory indicated by size1
Notify with flg. If the reference image number Idx exceeds the range represented by the short-term storage area size STsize and the long-term storage area size LTsize, refer to the calculation results of a plurality of images to set the extended motion compensation instruction flag AdvPred. Indicate. For example, when the reference image number Idx is assigned so that the number increases in the order of short-term storage area, long-term storage area, and extended motion compensation, the reference image number Idx is short-term storage area size STsize and long-term storage area size STsize. Storage area size LTsi
Judging by whether it exceeds the sum of ze.

(Embodiment 3) FIG. 6 is a block diagram of an image decoding apparatus using the image decoding method of the present invention.
The coded signal Str2 coded in the block diagram of the image coding apparatus using the image coding method of the present invention is decoded.
In the same figure, devices that perform the same operations as in the block diagram of the image encoding apparatus using the image encoding method of the present invention in FIG. 1 are assigned the same numbers and their explanations are omitted.

The variable length decoder VLD1 decodes the encoded signal Str2, and the picture type PTYPE, motion vector MV, modified reference image number RIdx, long-term storage area size LTsize, short-term storage area size STsize, block prediction type MBtype. ,
Outputs the quantized value DecQ.

The memory access controller MemAlloc4 uses the reference image candidate picture type StrPTY from the frame memory Mem4.
Obtaining PE, extracting the referenceable image from the images stored in the frame memory Mem4 by comparing it with the block prediction type MBtype, and obtaining the reference image number Idx corresponding to the unreferenceable image, The modified reference image number RId is obtained by performing the reverse operation of the memory access controller MemAlloc3 in the block diagram of the image encoding apparatus using the image encoding method of the present invention in FIG.
The reference image number Idx is restored from x. In addition, the reference image number
It is determined whether Idx indicates an area designated by the long-term storage area size LTsize and the short-term storage area size STsize. If the value is out of the range, the extended motion compensation instruction flag Ad
Indicate by vPred that the motion compensation is extended.

The operation of the inverse quantizer IQ and the inverse orthogonal transformer IT is shown in the figure.
1 is the same as the block diagram of an image encoding device using the image encoding method of the present invention of 1. The frame memory Mem2 outputs the reference image number Idx, the long-time storage instruction flag LTflg, the extended motion compensation instruction flag AdvPred, and the pixels in the area indicated by the motion vector MV as the motion compensation reference image Refpel, and the motion compensator MC performs the extended motion. The motion compensation image MCpel is output by performing the calculation indicated by the compensation instruction flag AdvPred (calculation using multiple images) and the block prediction type MBtype (average of both images).

The adder Add2 adds the decoded residual image DecDif and the motion-compensated image MCpel to output the decoded image signal Vout, and stores it in the frame memory Mem4 for reference in decoding the subsequent image. With the above configuration, the coded signal Str2 coded in the block diagram of the image coding apparatus using the image coding method of the present invention in FIG. 1 can be correctly decoded.

(Embodiment 4) FIG. 7 shows a memory access controller MemAlloc4 for implementing the image decoding method of the present invention.
It is a block diagram of. The block diagram of the memory access controller MemAlloc4 for realizing the image decoding method of the present invention in FIG. 7 is the same as the block diagram of the memory access controller MemAlloc3 for realizing the image encoding method of the present invention in FIG. Devices that operate are given the same numbers, and explanations thereof are omitted.

The short-time storage area size STsize and the long-time storage area size LTsize are input to the short-time storage buffer size calculation holder STbuf and the long-time storage buffer size calculation holder LTbuf, and the short-time storage area size STsize and Until the long-term storage area size LTsize is updated, the short-term storage area size STsize and the short-term storage area size STsize with the same value as the long-term storage area size LTsize
Output 1 and long-term storage area size LTsize1.

The unused buffer INDEX addition / correction device ModIdx2 determines whether or not the image in the area corresponding to the reference image number Idx is stored, and the reference image candidate picture type StrPTYPE
The corrected reference image number RIdx is not assigned to the area in which the image is not stored, and therefore the number for the area is added to the corrected reference image number RIdx. Also, block prediction type MBtype and reference image candidate picture type StrP
TYPE is compared to determine whether the image indicated by the reference image candidate picture type StrPTYPE can be referred to by the block prediction type MBtype. If it is not possible, the corrected reference image number RI
Since dx is not assigned, the number for that area is added to the corrected reference image number RIdx. In this way, FIG.
In the block diagram of the memory access controller MemAlloc3 for realizing the image encoding method of the present invention, the reference image number Idx is converted to the modified reference image number RIdx, and the inverse operation is performed to perform the original reference image from the modified reference image number RIdx. The number Idx is restored and output.

The corresponding buffer determiner BankSel1 compares the reference image number Idx with the short-term storage area size STsize and the long-term storage area size LTsize1, and the reference image number Idx is indicated by the short-term storage area size STsize. Long-term storage area size LT Long-term storage instruction flag LT indicating whether it is within the range of long-term storage memory indicated by size1
Notify with flg. If the reference image number Idx exceeds the range represented by the short-term storage area size STsize and the long-term storage area size LTsize, refer to the calculation results of a plurality of images to set the extended motion compensation instruction flag AdvPred. Indicate. For example, if the reference image number Idx is assigned so that it increases in the order of short-time storage area, long-time storage area, and extended motion compensation, the reference image number Idx is short-time storage area size STsize and long-time storage area size STsize. Storage area size LTsi
Judging by whether it exceeds the sum of ze.

As described above, the block diagram of the memory access controller MemAlloc4 for realizing the image decoding method of the present invention can be realized.

(Fifth Embodiment) Further, a program for realizing the configurations of the image coding method and the image decoding method shown in each of the above embodiments is recorded in a storage medium such as a flexible disk. This makes it possible to easily carry out the processing shown in each of the above-described embodiments in an independent computer system.

FIG. 8 shows the above-described first to fourth embodiments.
It is explanatory drawing in the case of implementing by a computer system using the flexible disk which stored the image coding method of 14 and the image decoding method.

FIG. 8B shows the appearance of the flexible disk as viewed from the front, the cross-sectional structure, and the flexible disk, and FIG. 8A shows an example of the physical format of the flexible disk that is the recording medium body. . The flexible disk FD is built in the case F, and a plurality of tracks Tr are formed concentrically from the outer circumference toward the inner circumference on the surface of the flexible disk FD, and each track is 1 in the angular direction.
It is divided into 6 sectors Se. Therefore, in the flexible disk storing the program, the image encoding method and the image decoding method as the program are recorded in the area allocated on the flexible disk FD.

FIG. 8 (c) shows a structure for recording / reproducing the program on / from the flexible disk FD. When recording the program on the flexible disk FD, the image encoding method and the image decoding method as the program are written from the computer system Cs via the flexible disk drive. Also,
When the image encoding method and the image decoding method are constructed in the computer system by the program in the flexible disk, the program is read out from the flexible disk by the flexible disk drive,
Transfer to computer system.

In the above description, a flexible disk is used as the recording medium, but an optical disk may be used as well. Moreover, the recording medium is not limited to this, and any recording medium such as a CD-ROM, a memory card, and a ROM cassette can be used as long as the program can be recorded.

[0060]

As described above, according to the image coding method and the image decoding method of the present invention, the reference image number Idx
By more efficiently encoding, the compression rate can be increased as compared with the conventional image encoding method and image decoding method, and its practical value is high.

[Brief description of drawings]

FIG. 1 is a block diagram of an image encoding device using an image encoding method of the present invention (Embodiment 1)

FIG. 2 is an explanatory diagram of a decoded image management method when B pictures are not stored in the frame memory (Embodiment 1)

[Fig. 3] P when a B picture is stored in a frame memory
Explanatory diagram of a decoded image management method for picture encoding (Embodiment 1)

[Fig. 4] B when a B picture is stored in a frame memory
Explanatory diagram of a decoded image management method for picture encoding (Embodiment 1)

FIG. 5 is a block diagram of a memory access controller MemAlloc3 for implementing the image encoding method of the present invention (embodiment).
2)

FIG. 6 is a block diagram of an image decoding apparatus using the image decoding method of the present invention (Embodiment 3)

FIG. 7 is a block diagram of a memory access controller MemAlloc4 for implementing the image decoding method of the present invention (Embodiment
Four)

FIG. 8 is an explanatory diagram of a storage medium for storing a program for realizing the image coding method and the image decoding method of the first to fourth embodiments by a computer system (fifth embodiment).

FIG. 9 is a block diagram of an image coding apparatus using a conventional image coding method.

FIG. 10 is an explanatory diagram of image coding using a short-time storage memory.

FIG. 11 is an explanatory diagram of image encoding using a long-term storage memory.

FIG. 12 is an explanatory diagram of a conventional decoded image management method.

FIG. 13 is a block diagram of a memory access controller MemAlloc1 for realizing a conventional image encoding method.

FIG. 14 is a block diagram of an image decoding apparatus using a conventional image decoding method.

[Explanation of symbols]

Vin Screen signal Vout Decoded image signal Str1, Str2 Coded signal T Orthogonal transformer IT Inverse orthogonal transformer Q Quantizer IQ Dequantizer ME Motion detector MC Motion compensator Mem1, Mem2, Mem3, Mem4 Frame memory MemAlloc1, MemAlloc3, MemAlloc4 Memory access controller VLC, VLC1 Variable length encoder VLD, VLD1 Variable length decoder Cs Computer system FD Flexible disk FDD Flexible disk drive

   ─────────────────────────────────────────────────── ─── Continued front page    (72) Inventor Toshishi Kondo             1006 Kadoma, Kadoma-shi, Osaka Matsushita Electric             Sangyo Co., Ltd. F-term (reference) 5C059 MA05 MA21 MC11 ME01 NN01                       SS20 UA02 UA05 UA33 UA34                       UA36 UA38 UA39                 5J064 AA01 BA09 BA13 BB03 BC01                       BC08 BC14 BC16 BD03

Claims (14)

[Claims] 1. An image encoding method for encoding by referring to a predetermined number of images selected from a plurality of reference images, wherein a buffer size indicating a buffer area for storing a reference image is determined, and the determined buffer size is determined. If the index indicating the image to be referred to indicates an image stored in the buffer area, the image is referred to, and if the index indicating the image to be referred indicates outside the buffer area, the inside of the buffer area is encoded. An image coding method for coding with reference to a result of an operation performed on a plurality of images stored in. 2. The image coding method according to claim 1, wherein the buffer area has a long-term storage area and a short-term storage area. 3. The image coding method according to claim 2, wherein the buffer size represents a long-term storage area and a short-term storage area. 4. The image coding method according to claim 1, wherein an index is not assigned to an area in which an image is not stored in the buffer area. 5. The image encoding method according to claim 1, wherein an index is not assigned to an area in the buffer area where an image that cannot be referred to by the encoding is stored. 6. An image decoding method for decoding by referring to a predetermined number of images selected from a plurality of reference images, wherein a buffer size indicating a buffer area for storing a reference image is decoded, When the index shown indicates an image stored in the buffer area, the image shown is referred to, and when the index indicating the image to be referred indicates outside the buffer area, the plurality of images stored in the buffer area are displayed. An image decoding method for decoding by referring to the result of calculation. 7. The image decoding method according to claim 6, wherein the buffer area has a long-term storage area and a short-term storage area. 8. The image decoding method according to claim 7, wherein the buffer size represents a long-term storage area and a short-term storage area. 9. The image decoding method according to claim 6, wherein no index is assigned to an area in which no image is stored in the buffer area. 10. The image decoding method according to claim 6, wherein an index is not assigned to an area in the buffer area in which an image that cannot be referred to by the encoding is stored. 11. An image coding apparatus for coding by referring to a predetermined number of images selected from a plurality of reference images, comprising buffer means for storing reference images, and a buffer size indicating a storage area of the buffer means. , A buffer size determining means for determining the determined buffer size, a buffer size encoding means for encoding the determined buffer size, and when the index indicating the image to be referred to indicates the image stored in the buffer area, the indicated image is referred to. If the index indicating the image to be referred indicates outside the buffer area, the image provided with an image encoding means for encoding with reference to the result of the operation on the plurality of images stored in the buffer area Encoding device. 12. An image decoding apparatus for decoding by referring to a predetermined number of images selected from a plurality of reference images, comprising buffer means for storing the reference image and buffer size indicating a storage area of the buffer means. And a buffer size decoding means for decoding the image to be referred to, if the index indicating the image to be referred to indicates the image stored in the buffer area, the image is referred to, and the index indicating the image to be referred to is outside the buffer area. In the case shown, the image decoding device is provided with an image decoding means for decoding by referring to the result of the operation on the plurality of images stored in the buffer area. 13. A storage medium for storing a program for executing the image coding method according to claim 1 by a computer, wherein the program refers to a computer a predetermined number of images selected from a plurality of reference images. An image encoding method for encoding by determining a buffer size indicating a buffer area for storing a reference image, encoding the determined buffer size, and storing an index indicating an image to be referenced in the buffer area. When the index indicating the referred image indicates outside the buffer area, the result obtained by performing the operation on the plurality of images stored in the buffer area is referred to. A storage medium for performing an image encoding method of encoding by means of a method. 14. A storage medium for storing a program for executing the image decoding method according to claim 6 by a computer, wherein the program refers to a computer a predetermined number of images selected from a plurality of reference images. An image decoding method for decoding by decoding a buffer size indicating a buffer area for storing a reference image, and an index indicating an image to be referenced indicates an image stored in the buffer area. An image decoding method that refers to an image and, when an index indicating the image to be referred indicates outside the buffer area, refers to a result of an operation performed on a plurality of images stored in the buffer area to perform decoding A storage medium characterized in that