JP2002027483A - Picture coding system, picture decoding system, and storage media - Google Patents

Abstract

PROBLEM TO BE SOLVED: To decrease propagation of data error caused by transmission error and the like in a picture coding using both intra-frame predictive coding and inter-frame predictive coding. SOLUTION: When there is a slice applied by the intra-frame predictive coding in the current frame stored in a buffer a difference detection circuit 2, checks the finite difference between the slice and a slice in the same position in the immediately previous frame stored by itself and notifies a control section 3 of it. For the large finite difference, the control section 3 forcedly applies the intra-frame predictive coding to the slice in the same position in the following frame of the current frame regardless the difference of the reference picture considered by motion compensation notified by a predictive circuit 108 when the following frame becomes the current frame.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a technique for reducing image quality degradation caused by a data error in an image coding system using both intra-frame prediction coding and inter-frame prediction coding.

[0002]

2. Description of the Related Art Conventionally, the MPEG2 system (ISO / IEC 13818-2) has been known as a technique for encoding a moving image by using both intra-frame prediction coding and inter-frame prediction coding.

[0003] Here, the intra-frame predictive coding is a technique for compressing and coding an image in each frame using only the spatial redundancy of the image in one frame.
The inter-frame predictive coding is a technique for compressing and coding an image in each frame using a correlation with another frame. Note that a frame refers to individual still images that make up a moving image. A moving image is expressed by, for example, sequentially displaying 30 frames per second.

Generally, in a moving image representing a series of scenes,
Since there are more stationary areas than moving areas, the correlation between frames is stronger, and inter-frame predictive coding can compress images more efficiently than intra-frame predictive coding. it can.

[0005] In the MPEG2 system, motion-compensated inter-frame prediction coding is performed to reduce redundancy in the time axis direction, and high compression efficiency is obtained. Motion-compensated inter-frame predictive coding means that a frame to be coded is 16 × 16
The macroblock is divided into macroblocks composed of pixels, and for each macroblock, a macroblock closest to the reference image is set as a comparison target macroblock, a difference image from the comparison target macroblock is obtained, and this is encoded. .

In the MPEG2 system, a macroblock coded by such motion-compensated inter-frame predictive coding is called a non-intra macroblock, and is coded by intra-frame predictive coding, not by motion-compensated inter-frame predictive coding. The resulting macro block is called an intra macro block. A picture composed of only intra macroblocks is called an I picture, a picture that uses only image data of a past picture in display order as reference image data is called a P picture, and is referred to as reference image data in display order. A picture in which image data of both past and future pictures can be used is called a B picture. It should be noted that only I-picture and P-picture image data are used as reference image data.

When a data error occurs in the motion-compensated inter-frame prediction, the influence of the error propagates to subsequent frames one after another during decoding. Therefore, MPEG2
In the method, a frame is encoded as an I picture at a predetermined cycle, for example, every 12 frames.

[0008] In the MPEG2 system, variable length coding is employed, in which values having a higher frequency of occurrence are represented by shorter codeword lengths. According to such variable length coding, the entire codeword length can be minimized.

By the way, if a data error occurs in the variable-length coding, a break between the code word and the next code word cannot be recognized, and the effect propagates to subsequent frames one after another. Therefore, in the MPEG2 system, an image head signal (PSC: Picture Start Code) indicating the head of an image signal and the head of a slice (composed of a plurality of macroblocks) obtained by dividing the screen into a horizontally long rectangle are used. The indicated slice head signal (slice header) is inserted at a position corresponding to the frame head or the slice head in the variable-length codeword string. As a result, even if a data error occurs, the variable-length codeword can be correctly detected from the beginning of the next frame or slice, thereby preventing the propagation of the data error. In the MPEG2 system, a bit string in which 23 bit values “0” continue and a bit value “1” is used as each head signal is used. In a general code word, a bit sequence between a code word and a next code word is used. When the bit value “0” is continuous, the number of consecutive bits is always less than 23, so that each head signal can be reliably detected from the variable-length codeword string.

In the MPEG2 system, each frame is determined to be one of the above-described I picture, P picture, and B picture by generally n / m frame periods between I pictures arranged in n frame periods. Is performed by arranging a P picture and a B picture between the I picture and the P picture (n, m
Is a natural number constant). By doing so, the I-pictures are periodically arranged, so that the errors accumulated by the inter-frame prediction coding are periodically eliminated by the I-pictures.

By the way, a moving image to be encoded is analyzed,
There is also known a technique of dynamically determining, based on the analysis result, any of the above-described I picture, P picture, and B picture for each frame constituting the moving image. For example, Japanese Patent Application Laid-Open No. 8-65678 describes a technique for switching the constants n and m, that is, the periods of the I picture and the P picture according to the correlation between adjacent frames. Japanese Patent Application Laid-Open No. 2000-32474 describes a technique for estimating the effect of motion prediction and determining whether an input frame is a B picture or an I or P picture in accordance with the effect. ing.

[0012]

However, in the above-mentioned prior art, each frame constituting a moving image is stored in an MPEG format.
When transmitting data obtained by compression encoding using two methods, the possibility of propagation of a data error caused by a transmission error or the like cannot be completely excluded.

For example, when a data error occurs in an I-picture due to a transmission error or the like, according to the above-described conventional technique, the effect on a frame in which the data error has occurred is propagated one after another to the next I-picture. The image quality of the moving image will be greatly deteriorated.

In a system for transmitting encoded data, when a data error occurs due to a transmission error or the like, if the transmitting side and the receiving side cooperate with each other to perform processing such as retransmission of a frame, the data is transmitted. It is also possible to eliminate the deterioration of the image quality of a moving image due to the propagation of an error. However, such processing has a disadvantage such as a delay when transmitting a moving image in real time. Also, like this,
The process performed by the transmitting side and the receiving side in cooperation when a data error occurs due to a transmission error or the like is performed when the encoded data is recorded and then reproduced, that is, the transmitting process and the receiving This cannot be applied when the reproduction process on the side operates asynchronously.

The present invention has been made in view of the above circumstances, and an object of the present invention is to provide a method for transmitting a moving image in real time, or a method for asynchronously performing transmission processing on a transmission side and reproduction processing on a reception side. It is an object of the present invention to reduce the propagation of a data error due to a transmission error or the like even when it operates.

[0016]

In order to solve the above-mentioned problems, the present invention provides a method for dividing a whole frame or a plurality of frames into a plurality of frames constituting a moving image. With the area as the unit area,
Applying intra-frame predictive coding or inter-frame predictive coding to an image coding apparatus that codes the moving image, a unit area that may be referred to in the inter-frame predictive coding is obtained by intra-frame predictive coding. The first to be encoded
When the intra-region is set as the intra-region, a unit region located at the same position as the first intra-region in a frame next to the frame including the first intra-region can be referred to in the inter-frame predictive coding. Intra-region setting means for setting a second intra region to be coded by intra-frame predictive coding.

According to the present invention, when the first intra area is set in a frame, the unit area located at the same position as the first intra area in the next frame of the frame is the second intra area. It is set in the intra area. That is, each unit area at the same position in two consecutive frames is an intra area to which intra-frame prediction coding is applied, which may be referred to by others in inter-frame prediction coding.

Therefore, for example, when the unit area is the entire area of a frame, the decoding side includes,
If a data error occurs due to a transmission error or the like, a change is made so that a frame that uses this rear frame as a reference frame in the inter-frame prediction coding refers to the front frame of the two frames. The effect of a data error propagated to a frame can be reduced. In addition,
In the present invention, as described above, the unit regions at the same position in two consecutive frames are intra-frames that may be referred to by others in the first and second inter-frame predictive codings, respectively. Area. Therefore, when the reference region in the inter-frame predictive coding is changed from the second intra region to the first intra region, the effect is suppressed to the difference between two consecutive frames.
Normally, it can be expected that the contents of successive frames are similar, so that it is expected that the error due to the change of the reference area is small.

The intra-region setting means may include a first intra-region and a unit located at the same position as the first intra-region in a frame immediately before the frame including the first intra-region. When the absolute value of the difference from the area is larger than a predetermined threshold, the unit area located at the same position as the first intra area in a frame next to the frame including the first intra area is referred to as the first area. It is good to set it to 2 intra-areas.

On the decoding side, when a data error occurs due to a transmission error or the like in the first intra area, the data in the frame immediately before the frame including the first intra area is
It is also possible to substitute the decoding result for the unit area located at the same position as the first intra area as the decoding result for the first intra area. However, in this case, if the difference between the first intra area and the unit area located at the same position as the first intra area in the immediately preceding frame is large, an error corresponding to the difference will be made thereafter. Propagate.

According to the present invention, by performing the above, the difference between the first intra area and the unit area at the same position as the first intra area in the immediately preceding frame is reduced. If the size is larger, a unit area at the same position as the first intra area in a frame next to the frame including the first intra area is set as the second intra area. Therefore, even when a data error occurs due to a transmission error or the like in the first intra area, an error corresponding to a large difference between the first intra area and a unit area located at the same position in the immediately preceding frame. But will not be propagated.

[0022]

DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the present invention will be described below.

In this embodiment, in accordance with the above-described MPEG2 method, a moving picture composed of a plurality of image frames using an I picture to which intra-frame predictive coding is applied and a P picture to which inter-frame predictive coding is applied. A case where an image is encoded and decoded will be described as an example.

First, an image coding apparatus to which the present embodiment is applied will be described.

FIG. 1 shows a configuration of an image coding apparatus to which the present embodiment is applied.

In this figure, image frames (current image frames) sequentially output from the camera 101 are A / D
The data is digitized by the converter 102 and temporarily stored in the buffer 152.

The prediction circuit 108 currently has a buffer 152
, The image frame (reference image frame) immediately before the current image frame stored in the current image frame is held, and the motion estimation process of each macro block of the current image frame is performed using the reference image frame. Then, the sum total of the difference between each macroblock of the current image frame and each comparison target macroblock (motion-compensated macroblock) in the reference image frame held by itself is calculated for each slice of the current image frame. The control unit 103 notifies the control unit 103 of the result together with the motion amount of each macroblock.

In response to this, the control unit 103 determines that the slice of the current image frame whose sum of the differences is equal to or less than a predetermined threshold value (which is determined according to the required image quality, compression efficiency, etc.) The prediction circuit 108 is instructed to apply the inter-frame predictive coding and to apply the intra-frame predictive coding to a slice of the current image frame larger than the predetermined threshold.

In response to this, the prediction circuit 108 divides the slice of the current image frame from the buffer 152 into macroblock units for the slice of the current image frame notified of the application of the inter-frame prediction coding. At the same time as reading, the macro block to be compared with the macro block read from the buffer 152 (the motion-compensated macro block) is read from the reference image frame held therein. Both are input to the subtractor 151,
A signal representing the difference is input to converter 104. On the other hand, for the slice of the current image frame that has been notified that intra-frame prediction coding should be applied, the buffer 1
From 52, the relevant slice of the current image frame is read in macroblock units. At this time, the macro block to be compared is not read from the reference image frame held by itself. Therefore, a signal representing a macroblock read from the buffer 152 is directly input to the converter 104 via the subtractor 151.

The image signal output from the subtractor 151 as described above is subjected to a discrete cosine transform (D
CT) and quantized by the quantizer 105, and then encoded by the variable-length encoder 109 together with management information including the type and the amount of motion of the intra / non-intra macroblock. This coded image signal is input to the TS transmission generation circuit 110, stored in a TS packet, and multiplexed with other media (such as voice and data) input from a portion not shown. After the multiplexed TS packet is added with an error correction code by the error correction circuit 111,
The data is output from the transmission path interface 112 and transmitted to the image decoding device on the receiving side.

The signal quantized by the quantizer 105 is inversely quantized by the inverse quantizer 106, and then subjected to inverse discrete cosine transform by the inverse transformer 107. As a result, the signal output from the subtractor 151 (the difference signal between the macroblock and the comparison target macroblock for a non-intra macroblock, and the signal representing the macroblock for an intra macroblock) is restored. You. If the restored signal is for a non-intra macroblock, the prediction circuit 108 reads the comparison target macroblock from the reference image frame held by itself and outputs the read macroblock to the adder 153. This allows
Restore the signal representing the macroblock. Then, the prediction circuit 108 restores the image frame by storing the signal representing the macroblock restored in this way. Then, the image frame is held as a new reference image frame.

The above operation is basically the same as the operation of a normal image coding apparatus to which the MPEG2 system is applied.
In addition to the above, in the present embodiment, the difference detection circuit 2 is newly provided, and when intra-frame prediction coding is applied to a certain slice in the current image frame, the difference detection circuit 2
Causes the control unit 103 to calculate the difference between the current slice and the slice at the same position as the current slice in the image frame immediately before the current image frame. The intra-frame predictive coding is forcibly determined according to the difference for the slice at the same position as the slice.

That is, the difference detection circuit 2 includes the adder 15
Then, a signal representing the macroblock output from 3 is accumulated to restore the image frame, and this is held as a new reference image frame. Further, the difference detection circuit 2 includes a control unit 10
3 notifies the prediction circuit 108 of the type of prediction coding applied to each slice of the current image frame. If there is a slice to which intra-frame predictive coding is applied in the current image frame, the slice of the current image data is read from the buffer 152, and the slice is read at the same position in the reference image frame held by itself. Is obtained, and the difference is notified to the control unit 103. When the difference is notified, the above-mentioned monitoring may be stopped at least for the image frame next to the current image frame. In this way, in the processing by the control unit 103 described later, a situation in which slices at the same position frequently apply intra-frame predictive coding continuously over three or more image frames is avoided. Can be prevented.

Upon receiving the notification from the difference detection circuit 2, the control unit 103 sets the difference for a certain slice indicated in the notification to a predetermined threshold value (this is the influence of the data error in the image decoding device described later). Greater than the current image frame), if the next image frame is processed as the current image frame for a slice at the same position as that slice in the next image frame of the current image frame Irrespective of the difference notified from the prediction circuit 108, the prediction circuit 108 is forcibly notified to apply the intra-frame prediction coding.

With the above operation, in the image coding apparatus of the present embodiment, the slice of interest (slice of interest) and the slice of interest of the image frame (frame of interest) to which intra-frame predictive coding has been applied. If the difference between the slice of interest and the slice at the same position in the image frame immediately before the frame is large, the slice at the same position as the slice of interest in the next image frame of the frame of interest must be used. Even so, intra-frame prediction coding will be applied.

Next, an image decoding apparatus to which the present embodiment is applied will be described.

FIG. 2 shows a configuration of an image decoding apparatus to which the present embodiment is applied.

In the figure, an MPEG2 TS packet received from the transmission line interface 112 is subjected to error correction processing by an error correction circuit 113. The TS reception decoding circuit 114 separates the MPEG2 TS packet output from the error correction circuit 113 into a TS packet storing an encoded image signal and data of another medium (audio, data, etc.). Data of other media is output to a portion not shown. On the other hand, the TS packet
The S-reception decoding circuit 114 analyzes header information and the like of the packet, extracts an encoded image signal therefrom, and decodes the encoded image signal into management information and an image signal by the variable-length decoder 115. The management information is notified to the prediction circuit 118,
The image signal is inversely quantized by the inverse quantizer 116, and then inverse discrete cosine transformed by the inverse transformer 117. As a result, when the image signal is for an intra macroblock, a signal representing the macroblock is restored. In the case of a non-intra macro block, a signal representing a difference between the macro block and a macro block to be compared with the macro block (motion-compensated macro block) is restored.

The prediction circuit 118 includes a variable length decoder 115
Refer to the management information notified from. If the signal output from the inverse converter 117 is for a non-intra macro block, the macro block to be compared (the macro block whose motion is compensated) is read out from the reference image frame held by itself and output. I do. On the other hand, in the case of an intra macroblock, the comparison target macroblock is not output. As a result, a signal representing the macroblock is output from the adder 161.

The signal output from the adder 161 as described above is supplied to a D / A
After being input to the converter 119 and converted into an analog signal, it is output from the image output terminal 120. The signal output from the selector 162 is stored in the prediction circuit 118. The prediction circuit 118 restores an image frame based on the signal representing the macroblock thus accumulated, and holds this as a new reference image frame.

The above operation is basically the same as the operation of a normal image decoding apparatus to which the MPEG2 system is applied.
In addition to the above, in the present embodiment, the selector 162 is newly provided, and the prediction circuit 118 controls the selector 162 as follows based on the error correction result of the error correction circuit 113.

That is, the error correction circuit 113
If error correction for the G2 TS packet fails,
This is notified to the prediction circuit 118. In response to this, the prediction circuit 118 controls the selector 162 to switch the input from the adder 161 to the prediction circuit 118. On this occasion,
When the management information received from the variable-length decoder 115 indicates a non-intra macroblock for the coded image signal stored in the TS packet for which error correction has not been performed, as described above, It reads and outputs a comparison target macroblock determined by the amount of motion included in the management information from the reference image frame held by itself.

On the other hand, if the management information received from the variable length decoder 115 indicates an intra macroblock, error correction could not be performed based on the processing results up to now (for example, the number of processed macroblocks from the beginning of the slice). TS
The position of the macroblock in the frame obtained from the coded image signal stored in the packet is specified, and the macroblock at the position is read from the reference image frame held by itself and output.

According to the above operation, in the image decoding apparatus of the present embodiment, if a macroblock for which error correction was not possible is a non-intra macroblock, the image frame including the macroblock is replaced with the macroblock. A signal representing a motion-compensated macroblock in the immediately preceding image frame (reference image frame) is output from the selector 162. If the signal is an intra macroblock, the signal includes the macroblock instead of the macroblock. A signal representing a macroblock located at the same position in the image frame immediately before the image frame is output from the selector 162. Also, the prediction circuit 1
18 stores the image frame obtained by accumulating the signal output from the selector 162 as a reference image frame. For a macroblock for which error correction has failed, the macroblock is replaced with the macroblock. The macroblock located at the same position or the motion-compensated macroblock in the image frame immediately before the image frame including is held as the reference image frame.

Hereinabove, one embodiment of the present invention has been described.

In the present embodiment, on the encoding side, there is a slice to which intra-frame predictive coding is applied in the current image frame, and the slice and the image frame immediately before the current image frame (see the reference). If the difference between the slice at the same position as the slice in the image frame and the slice at the same position is larger than a predetermined threshold, the intra-frame prediction code Is enforced.
That is, intra-frame prediction coding is applied to slices at the same position in two consecutive frames.

In the present embodiment, on the decoding side, two slices at the same position of two consecutive frames to which intra-frame predictive coding has been applied are:
If the data error of the intra macro block included in the succeeding slice cannot be corrected, the non-intra macro block which is to be decoded by referring to the intra macro block is normally the intra macro block in the preceding slice. Decoding is performed with reference to an intra macroblock located at the same position as the block. Therefore, in such a case, it is possible to reduce the influence of a data error propagated to an image frame after an image frame including an intra macroblock in which a data error has occurred. In addition, usually, since it can be expected that the contents of successive frames are similar,
The effect of changing the reference target of non-intra blocks is
It can be expected to be smaller than the effect caused by data errors.

In this embodiment, when a data error occurs in a certain macroblock on the decoding side, the data error occurs in the macroblock of the frame immediately before the image frame including the macroblock. Used instead of macroblocks. Therefore, as described above, when a data error of a certain intra macro block cannot be corrected without applying intra-frame prediction coding to slices at the same position of two consecutive frames, By using a macroblock located at the same position as the intra macroblock in the image frame immediately before the image frame including the intra macroblock instead of the macroblock, the intra macroblock in which the data error has occurred can be used. It is possible to reduce the effect of a data error propagated to an image frame after the included image frame.

However, in this case, if the difference between the image frame including the intra macroblock in which the data error has occurred and the frame immediately before the image frame is large,
An error according to the difference propagates to subsequent frames. Therefore, in the present embodiment, as described above, the difference between the slice to which intra-frame prediction coding is applied and the slice located at the same position as the slice in the image frame immediately before the image frame including the slice is described. Is larger than a predetermined threshold, intra-frame predictive coding is forcibly applied to a slice at the same position as the slice in an image frame next to the image frame including the slice. By doing so, of the two slices at the same position of two consecutive frames to which intra-frame predictive coding has been applied, the preceding slice (the same as the previous frame from the frame including the slice) Slice with a large difference from the slice at the position)
As a general rule, reference is not made in inter-frame predictive coding unless a data error occurs in a subsequent slice. By doing so, it is possible to more effectively reduce the influence of an error when a data error occurs in the preceding slice.

The present invention is not limited to the above embodiment, and various modifications are possible within the scope of the invention.

For example, the above embodiment has been described by taking as an example a case where the coding side determines whether to apply intra-frame prediction coding or inter-frame prediction coding on a slice basis. However, it is needless to say that the present invention performs the above-described determination in a case where the image
Any area can be set as a unit.

In the above-described embodiment, the switching process of the selector 162 when the decoding side detects a data error in a macroblock is performed in units of macroblocks. However, the present invention can perform the above-described switching process in units of arbitrary regions, such as in the case of performing the process in units of image frames or in the case of performing units of slices. For example, when the above switching process is performed in slice units, in FIG.
A buffer that can store one slice of image signal is provided between the inverse transformer 117 and the adder 161. When the error correction circuit 113 notifies the image signal of one slice stored in the buffer of a data error for a certain macroblock, the prediction circuit 118 clears the contents of the buffer and The slice at the same position as the slice stored in the buffer may be read out from the reference image data held by itself, and output to the D / A converter 119 via the selector 162.

Further, in the above embodiment, a case has been described as an example in which a moving picture is coded / decoded using I pictures and P pictures in accordance with the MPEG2 system. The present invention can be widely applied to a case where a moving image is encoded / decoded in combination with inter-frame prediction encoding.

For example, in the MPEG2 system, when an I picture, a P picture and a B picture are arranged according to a predetermined routine as described in the section of the prior art,
Applicable. In this case, on the encoding side, the MPE
Prior to the encoding process according to the G2 system, the following process is performed to determine the arrangement of each picture.

That is, of the sequentially input frames, for a frame that is determined to be an I picture when the predetermined routine is performed, the frame immediately preceding the frame in the display order is stored in a separately prepared memory or the like. Check the difference between them. If the difference between the two is greater than a predetermined threshold, the frame immediately after the I-picture frame in the display order is also an I-picture,
Thereafter, the I picture, the P picture, and the B picture may be arranged again according to the predetermined routine. For example, when the predetermined routine is followed, each image frame constituting the moving image is represented by I1 → B1 ₁ → P1 ₁ → B1 ₂ → P
It is assumed to be assigned to each picture in _{1 2 → B1 3 → I2 →} B2 1 → ... order of. Here, if the difference between the image frame assigned to the I picture (I2) and the image frame assigned to the B picture (B1 ₃ ) is larger than a predetermined threshold, the image originally assigned to the B picture (B2 ₁ ) A frame is assigned to the I picture (I2 '),
After that, they are arranged again according to the predetermined routine. In this case, each image frame constituting the moving image is I1 → B1 ₁ → P1 ₁ → B1 ₂ → P1 ₂ → B1 ₃ → I2 → I2 ′ → B2 ₁ ′ →
Are assigned to each picture in the order of...

Further, the image encoding device and the image decoding device shown in the above-described embodiment are ASICs or FPGAs.
A hardware realized by an integrated logic IC such as the above may be used. Alternatively, a CPU, a memory, an external storage device such as a hard disk device, a reading device that reads data from a storage medium such as a CD-ROM, an input device such as a keyboard, an output device such as a display, and a communication device such as a network. Software implemented on a computer having a general configuration, including a communication device for communicating with other devices via a medium, and an interface for transmitting and receiving data between these devices. May be. In this case, a program for realizing the image encoding device and the image decoding device of the present embodiment as software is provided from a storage medium such as a CD-ROM via a reading device, or
After being installed in an external storage device from a communication medium such as a network via a communication device, it is loaded onto a memory from there and executed by the CPU.

[0057]

As described above, according to the present invention,
Even when it is desired to transmit a moving image in real time, or when the transmission process on the transmission side and the reproduction process on the reception side operate asynchronously, propagation of a data error due to a transmission error or the like can be reduced.

[Brief description of the drawings]

FIG. 1 is a schematic configuration diagram of an image encoding device to which an embodiment of the present invention has been applied.

FIG. 2 is a schematic configuration diagram of an image decoding device to which an embodiment of the present invention has been applied.

[Explanation of symbols]

101: camera, 102: A / D converter, 103:
Control unit, 104: Transformer, 105: Quantizer, 1
06, 116: inverse quantizer, 107, 117: inverse transformer, 108, 118: prediction circuit, 109: variable length encoder, 110: TS transmission generation circuit, 111, 113 ...
Error correction circuit, 112 ... transmission line interface, 1
14: TS reception decoding circuit 115: Variable length decoder
119 ... D / A converter, 120 ... Screen output terminal, 1
51: subtractor, 152: buffer, 161: adder, 162: selector

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Yuji Araki 216 Totsuka-cho, Totsuka-ku, Yokohama-shi, Kanagawa Prefecture Within the IP Systems Division of Hitachi, Ltd. (72) Inventor Susumu Takase 216 Totsuka-cho, Totsuka-ku, Yokohama-shi, Kanagawa F-term in the IP Systems Division of Hitachi, Ltd. (Reference) 5C059 MA00 MA04 MA05 MA23 ME01 NN01 NN26 NN28 NN31 PP04 PP05 RB02 RF01 RF05 RF14 SS20 TA25 TA76 TB06 TC02 TC11 TC31 TD05 TD06 TD12 UA34 5J064 AA03

Claims (6)

[Claims] 1. For each of a plurality of frames constituting a moving image, an entire region of the frame or each region obtained by dividing the frame into a plurality is defined as a unit region.
An image encoding apparatus that encodes the moving image by applying intra-frame predictive coding or inter-frame predictive coding, wherein a unit area that may be referred to in inter-frame predictive coding is an intra-frame predictive code. Is set in the first intra region encoded by the encoding, the first intra region in the frame next to the frame including the first intra region is set.
An intra-area setting unit that sets a unit area at the same position as the intra-area of the second intra-area encoded by the intra-frame predictive coding, which may be referred to in the inter-frame predictive coding. An image encoding device, comprising: 2. The image encoding apparatus according to claim 1, wherein the intra-region setting means includes a first intra-region and a frame in a frame immediately before a frame including the first intra-region. If the difference between the first intra region and the unit region located at the same position is larger than a predetermined threshold, the difference between the first intra region and the unit region located at the same position as the first intra region in a frame subsequent to the frame including the first intra region The image coding apparatus according to claim 1, wherein the second intra area is set for a unit area of the image encoding apparatus. 3. An image decoding device for decoding a moving image composed of a plurality of frames, which is encoded by the image encoding device according to claim 1 or 2, wherein a data error is detected for each unit area. Error detecting means for detecting, when a data error is detected in the unit area set in the second intra area by the error detecting means, the second intra area is to be referred to in inter-frame predictive coding. Reference area changing means for changing the unit area to refer to the first intra area set in the frame immediately before the frame including the second intra area. An image decoding device characterized by the following. 4. The image decoding apparatus according to claim 3, wherein when the error detection means detects a data error in a unit area set in the first or second intra area, the error of the area is determined. An image decoding apparatus, further comprising: a replacing unit that substitutes, as a decoding result, a decoding result of a unit area located at the same position as the area in a frame immediately before a frame including the area. 5. For each of a plurality of frames constituting a moving image, an entire region of the frame or each region obtained by dividing the frame into a plurality is defined as a unit region.
A storage medium storing an encoding program for applying the intra-frame prediction encoding or the inter-frame prediction encoding and encoding the moving image, wherein the encoding program is read by an electronic computer and executed. By doing so, if a unit area that may be referred to in inter-frame predictive coding is set as a first intra area to be coded by intra-frame predictive coding, the first intra area is The first frame in a frame next to the containing frame.
An intra-area setting unit that sets a unit area at the same position as the intra-area of the second intra-area encoded by the intra-frame predictive coding, which may be referred to in the inter-frame predictive coding. And a storage medium constructed on the computer. 6. A storage medium storing a decoding program for decoding a moving image composed of a plurality of frames, encoded by the encoding program stored in the storage medium according to claim 5. The decoding program is read and executed by an electronic computer, and configured so that an error detection unit that detects a data error for each unit area and the second intra area are set by the error detection unit. When a data error is detected in the unit area, the unit area scheduled to refer to the second intra area in the inter-frame prediction coding is set in a frame one frame before the frame including the second intra area. And a reference area change unit configured to change the first intra area so that the first intra area is referred to, on the computer.