JP2011130500A - Expansion decoding device, and expansion decoding method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To accelerate expansion decoding processing of a moving image, and to prevent increase in the device scale. <P>SOLUTION: An expansion decoding device is used for subjecting moving image data imaged by a plurality of cameras and subjected to compression encoding to expansion decoding. The expansion decoding device includes a determination means for determining the point of view, where a moving image at the point of view used for motion prediction is imaged in compression encoding for moving image data at a point of view of a specific camera to be subjected to expansion decoding, based on information showing whether the motion prediction using moving image data imaged by a camera at another point of view is executed, in relation to moving images imaged by the cameras at the respective point of view, when subjecting a moving image at a point of view of a specific camera to expansion decoding, and to determine the moving image data imaged by the camera at the point of view as being the moving image data used for the motion prediction; and an expansion decoding means to execute expansion decoding for the moving image data at the point of view determined by the determination means and moving image data imaged by the camera at the specific point of view to be subjected to the expansion decoding. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a technique that is effective when applied to an apparatus or a method for encoding / decoding an image captured from multiple viewpoints.

  In recent years, attention has been paid to a technique for utilizing moving images simultaneously captured from multiple viewpoints. By using such a moving image, it becomes possible that was impossible with the conventional stereo camera system. For example, it has become possible for a user to view a camera moving image from a free viewpoint without using a stereoscopic display. Specifically, by simultaneously capturing images of a concert pattern from multiple viewpoints, the user can view the concert situation not only from one viewpoint, but also from any viewpoint such as the horizontal direction or the backward direction. .

  By the way, the amount of information of moving images is generally very large. For this reason, performing media accumulation and network transmission without compressing moving image data is disadvantageous in terms of transmission speed and cost. For this reason, techniques for compressing and encoding moving image data in a reversible or irreversible manner have been developed. For example, MPEG-1, MPEG-2, MPEG-4, and the like standardized by the Moving Picture Experts Group (MPEG).

  However, as the number of viewpoints that simultaneously capture images increases (the number of cameras increases), the number of moving images also increases. For this reason, the total amount of data of moving images simultaneously captured from multiple viewpoints is larger than the amount of data of moving images captured using a single camera. For this reason, there is a demand for a technique for efficiently compressing and encoding moving image data captured simultaneously from multiple viewpoints.

  In order to solve such a problem, a technique for improving the prediction efficiency by using the correlation between moving images captured at each viewpoint has been proposed. In these techniques, the encoding efficiency is improved as the prediction efficiency is improved. The correlation between moving images taken at each viewpoint indicates that a part of an object or background reflected in a certain viewpoint camera is also reflected in a camera at another viewpoint. For example, when a frame in one viewpoint camera is compared with a frame at the same time in another viewpoint camera, the same object or the same background may be captured when both cameras are capturing images in close directions. . For this reason, it is possible to perform predictive coding using a motion vector by regarding frames captured at the same time by such different cameras as frames captured by the same camera. For example, the frame from the other camera (the former camera) is compared with the case where the frame of the latter camera is motion-predictively encoded using only the frame imaged by the same camera (the latter camera) as before. When the motion prediction encoding is further performed using, encoding efficiency can be improved. In this case, the motion vector corresponds to the parallax between the two cameras. Examples of such techniques include Patent Documents 1-7.

JP 2001-186516 A Special Table 2002-52394 JP 2002-300607 A Japanese Patent No. 3426668 Japanese Patent Laid-Open No. 06-98312 Japanese Patent Laid-Open No. 10-191394 JP 2000-23918 A

  The moving image data created by the technique as described above includes moving images of a plurality of viewpoints in one moving image data. For this reason, when decompression decoding is executed by the conventional decompression decoding technique, not only the moving image at the viewpoint desired by the user but also the moving image data at all other viewpoints included in the moving image data is expanded and decoded. It was converted. For this reason, processing capacity and processing time are spent on unnecessary processing, and problems such as processing delays occur.

  Therefore, the present invention solves these problems, and by not performing decompression decoding processing that is not necessary for decompression decoding of a moving image of a viewpoint desired by the user, the processing speed and device scale are increased. It is an object of the present invention to provide an apparatus, a method, and the like that can prevent the increase in the amount.

In order to solve the above problems, the present invention has the following configuration. One embodiment of the present invention provides:
A decompression decoding device that decompresses and decodes moving image data captured and compressed by a plurality of cameras,
When decompressing and decoding moving images from a specific camera viewpoint,
Based on the information indicating whether or not the motion prediction using the moving image data captured by the camera of the other viewpoint is performed on the moving image captured by the camera of each viewpoint, the specific camera to be decompressed and decoded A viewpoint that captures the viewpoint moving image used for motion prediction at the time of compression encoding is determined for the viewpoint moving image data, and the moving image data captured by the camera of the viewpoint is used as the moving image data used for the motion prediction. A judging means for judging;
Decompression decoding means for performing decompression decoding on the moving image data of the viewpoint determined by the determination means and the moving image data captured by the specific viewpoint camera to be decompressed and decoded;
Is a decompression decoding apparatus.

  One embodiment of the present invention may be realized by a program being executed by an information processing device. That is, the present invention can specify the processing executed by each unit in the above-described mode as a program for causing the information processing apparatus to execute or a recording medium on which the program is recorded. Further, the present invention may be specified by a method in which the information processing apparatus executes the processing executed by each of the above-described means.

  According to the present invention, it is possible to realize speeding up of processing and prevention of increase in apparatus scale when decompressing moving image data including moving images of a plurality of viewpoints.

It is a figure which shows the prediction encoding method of each frame imaged with the reference | standard camera, and the prediction encoding method of each frame imaged with the reference camera. It is a figure which shows the structural example of moving image data. It is a figure which shows the example of the functional block of a compression encoding apparatus. It is a flowchart which shows the process example of a reference | standard camera determination process and a reference destination determination process. It is a figure which shows the example of arrangement | positioning of a camera. It is a table | surface which shows whether each camera becomes a reference | standard camera and which reference | standard camera becomes a frame used as a reference place when it is a reference camera. It is a figure which shows the example of the functional block of an expansion | extension decoding apparatus. It is a flowchart which shows the process example of a decoding camera determination process. It is a figure which shows the trouble of the prior art. It is a figure which shows the trouble of the prior art.

[principle]
First, the principle of the predictive coding method according to the present invention will be described. FIG. 1 is a diagram illustrating a predictive encoding method for each frame imaged by the reference camera and a predictive encoding method for each frame imaged by the reference camera. In FIG. 1, the frame located at the root of the arrow is predictively encoded by referring to the frame located at the tip of the arrow.

  The I frame indicates a frame on which intraframe coding is performed. I frames are encoded without reference to any other frame. P frame indicates a forward temporal prediction encoded frame. The P frame is a frame captured by the same camera, and is predictively encoded by referring to the I frame or another P frame located immediately before in time. Such predictive encoding by referring to another frame imaged by the same camera is hereinafter referred to as “motion predictive encoding”. A B frame indicates a bidirectional temporal prediction encoded frame. A B frame is a frame captured by the same camera and is predictively encoded by referring to an I frame or P frame located immediately before in time and an I frame or P frame located immediately after in time. . The concepts of I frame, P frame, and B frame are the same as those of MPEG-1, 2, and 4.

  The I 'frame is predictively encoded by referring only to the I frame at the same time imaged by the reference camera. Such predictive encoding by referring to a frame imaged by another camera is hereinafter referred to as “parallax predictive encoding”. The P ′ frame is predicted by referring to the P frame at the same time imaged by the reference camera and the I ′ frame or other P ′ frame that is imaged by the same camera and positioned immediately before in time. Encoded. The B ′ frame is a B frame at the same time imaged by the reference camera, an I ′ frame or a P ′ frame imaged by the same camera and positioned immediately before in time, and a frame imaged by the same camera. The prediction encoding is performed by referring to the I ′ frame or the P ′ frame located immediately after in time.

  Next, the configuration of moving image data created by predictive coding according to the present invention will be described. FIG. 2 is a diagram illustrating a configuration example of moving image data. The moving image data includes SEQH and GOP. SEQH is inserted in all camera units of GOP (Group of pictures). SEQH (n) indicates that the following GOP is the nth GOP. The SEQH includes the total number of cameras, the parallax prediction type of each camera (that is, whether the camera that captured each frame is the base camera or the reference camera), and an identifier indicating the reference destination camera in each reference camera. GOP is data obtained by grouping frames arranged along the time axis. GOP (m, n) indicates the nth GOP of the camera m.

  One GOP includes GOPH and a plurality of frames. GOPH is GOP header information. The GOPH includes an identifier indicating the camera that has captured the frame, and time information during which no parallax prediction is performed (frame number relative value from the GOP head). Frame is encoded frame data.

One Frame includes a Frame Header and a plurality of MBs (Macro Blo
ck). Frame Header is header information of the frame. The Frame Header includes the prediction type (I, P, B, I ′, P ′, B ′) of the frame. MB indicates macroblock information.

  Each MB includes MBType, MV, and DCTCoeff. MBType includes the prediction type (Intra, Inter, Bi-Direction) and quantization coefficient of each macroblock. The prediction type further includes an identifier of a frame referred to in unidirectional prediction (Inter) or bidirectional prediction (Bi-Direction). The prediction type includes one identifier for unidirectional prediction and two identifiers for bidirectional prediction. MV is vector information. Hereinafter, this vector information is described by distinguishing it into a motion vector and a disparity vector. The motion vector indicates vector information between frames captured by the same camera, and the vector information between frames captured by a camera different from the disparity vector. DCTCoeff is quantization DCT coefficient information of the prediction error.

  Next, the compression encoding device 1 that generates the moving image data as described above by compressing and encoding the moving image captured by each camera, and decoding the moving image data generated by the compression encoding device 1 The decompression decoding apparatus 2 to be converted will be described.

[Compression encoding device]
First, a configuration example of the compression encoding device 1 will be described. The compression encoding apparatus 1 includes a CPU (Central Processing Unit), a main storage device (RAM), an auxiliary storage device, and the like connected via a bus in terms of hardware. The auxiliary storage device is configured using a nonvolatile storage device. The non-volatile memory device referred to here is a so-called ROM (Read-Only Memory: EPROM (Erasable Programmable Read-Only Memory), EEPROM (Electrically Erasable Programmable Read-Only Memory, etc.). ), Hard disk, etc.

  FIG. 3 is a diagram illustrating an example of functional blocks of the compression encoding device 1. The compression encoding device 1 loads various programs (OS, applications, etc.) stored in the auxiliary storage device into the main storage device and is executed by the CPU, whereby the input frame buffer 101, the subtractor 102, and DCT quantization are performed. It functions as a device including a unit 103, an IDCT inverse quantization unit 104, an adder 105, a frame buffer 106, a motion / disparity vector compensation unit 107, a variable length coding unit 108, a header addition unit 109, a control unit 110, and the like. DCT quantization section 103, IDCT inverse quantization section 104, motion / disparity vector compensation section 107, variable length coding section 108, header addition section 109, and control section 110 are realized by the program being executed by the CPU. . Further, the DCT quantization unit 103, the IDCT inverse quantization unit 104, the motion / disparity vector compensation unit 107, the variable length coding unit 108, the header addition unit 109, and the control unit 110 may be configured as a dedicated chip. Next, each functional unit included in the compression encoding device 1 will be described.

<Input frame buffer>
The input frame buffer 101 buffers a moving image input to the compression encoding device 1. The compression encoding apparatus 1 receives a moving image captured by the basic camera and a moving image captured by the reference camera. Therefore, the input frame buffer 101 buffers the moving image captured by the basic camera and the moving image captured by the reference camera. The input frame buffer 101 outputs frame data in units of encoding processing of each camera according to instructions from the control unit 110. The encoding processing unit may be one frame or a plurality of frames such as 1 GOP (Group of Pictures). Hereinafter, a frame output by the input frame buffer 101, that is, a frame to be subjected to compression encoding processing is referred to as an input frame.

<Subtractor>
The subtractor 102 calculates a difference between the input frame and prediction information by motion compensation or parallax compensation, and outputs the result as prediction difference information.

<DCT quantization unit>
The DCT quantization unit 103 performs a DCT (Discrete Cosine Transform) operation and a quantization operation. The DCT quantization unit 103 performs a DCT operation on the prediction difference information calculated by the subtractor 102 in units of blocks, quantizes the DCT coefficient, and outputs the resulting quantized DCT coefficient.

<IDCT inverse quantization unit>
The IDCT inverse quantization unit 104 performs an IDCT (Inverse Discrete Cosine Transform) operation (hereinafter also referred to as “inverse DCT operation”) and an inverse quantization operation. The IDCT inverse quantization unit 104 performs inverse quantization and inverse DCT operation on the quantized DCT coefficient to obtain an inverse DCT operation result.

<Adder>
The adder 105 adds the inverse DCT calculation result and the prediction result by motion compensation or parallax compensation to generate a local decoded moving image.

<Frame buffer>
The frame buffer 106 stores local decoded moving images. Further, the frame buffer 106 outputs a designated frame of the designated camera moving image according to an instruction from the control unit 110. Hereinafter, a frame output by the frame buffer 106, that is, a frame used for motion prediction or parallax prediction is referred to as a prediction source frame.

<Motion / parallax vector compensation unit>
The motion / disparity vector compensation unit 107 performs prediction by block matching using the input frame and the prediction source frame in accordance with an instruction from the control unit 110. The motion / disparity vector compensation unit 107 outputs the motion vector information and the disparity vector information to the variable length coding unit 108. In addition, the motion / disparity vector compensation unit 107 outputs the prediction information to the subtracter 102. Motion vector information, disparity vector information, and prediction information are used to minimize the prediction error. Also, the motion / disparity vector compensation unit 107 outputs motion vector information for the entire frame to the control unit 110. The motion vector information is, for example, the average and variance of motion vector information in all blocks.

<Variable length code part>
The variable length coding unit 108 generates compression-coded frame data by performing variable length coding on the quantization result. Also, the motion vector information used for motion compensation and the disparity vector information used for disparity compensation are passed to the header adding unit 109.

<Header addition part>
The header addition unit 109 adds information such as camera number, camera type (base camera or reference camera), other cameras to be referred to, etc. to each frame or a plurality of frames after compression encoding, thereby moving image data. Is generated.

<Control part>
The control unit 110 controls encoding of moving images captured by each camera and determines a reference camera (
A reference camera determination process), a frame to be referred to for predictive encoding of a frame imaged by the reference camera (reference destination determination process), and the like. Further, the control unit 110 is connected so as to be able to instruct each function unit. In addition, parameters of each camera (arrangement information about each camera) are input to the control unit 110 from the outside. The arrangement information may be an absolute position or a relative position of each camera). Hereinafter, the reference camera determination process and the reference destination determination process will be described.

  FIG. 4 is a flowchart illustrating an operation example of the control unit 110 in the base camera determination process and the reference destination determination process. An operation example of the control unit 110 will be described with reference to FIG. The following processing is executed for each set of GOPs. That is, the process of FIG. 4 is executed in units of a plurality of GOPs grouped by one SEQH.

  First, the control unit 110 determines whether panning has occurred in the last frame of the immediately preceding GOP (S01). For example, the control unit 110 determines the occurrence of panning based on motion vector information in the last frame of the immediately preceding GOP (for example, average and variance of motion vector information in all blocks). In other words, the occurrence of panning is determined based on the movement of the subject imaged in the last frame. In this case, the control unit 110 determines whether or not the average value of the horizontal motion vectors is equal to or greater than a threshold value and whether or not the variance is equal to or less than the threshold value. When these two conditions are satisfied, the control unit 110 determines that panning has occurred. On the other hand, when one of these two conditions is not satisfied, the control unit 110 determines that panning has not occurred.

  When it is determined that panning has not occurred (S01-No), the control unit 110 calculates C (N / 2 ± nK) and sets the camera corresponding to the calculation result as the reference camera (S03). C (m) is an identifier indicating the m-th camera, and it is assumed that numbers are assigned in the order in which the cameras are arranged in one direction. N indicates the total number of cameras. N represents an integer of 0 or more. Further, the value of K is a value indicating a positive value, and is a value that is appropriately set by the designer in accordance with the distance between the cameras, the distance between the camera and the subject, and the like. In this case, the center of the camera row and the cameras in the left and right directions at equal intervals (K) from the center are set as the reference cameras.

  Then, a reference frame for predictively encoding a frame captured by the reference camera among the frames included in the GOP to be processed is determined. In this case, the control unit 110 determines for each frame that the nearest base frame on the center side is the reference destination frame (S06). FIG. 5 is a diagram illustrating an arrangement example of cameras. FIG. 6 is a table showing whether or not each camera is a reference camera, and which reference camera is a reference frame when it is a reference camera. In FIG. 5, seven cameras C (1) to C (7) are arranged in numerical order on the X axis. In FIG. 5, the cameras are arranged at equal intervals or at arbitrary intervals in the direction perpendicular to the shooting direction (Z-axis direction). In FIG. 6, the value of K is set to “2”. In FIG. 6, “◯” indicates a reference camera, and “C (m)” indicates that a frame captured by the camera is a reference frame. When it is determined that panning has not occurred as described above, C (2), C (4), and C (6) are set as reference cameras as a result of the process of S03. C (1) is referred to as C (2), C (3) and C (5) are referred to as C (4), and C (7) is referred to as C (6).

Next, a case where it is determined that panning has occurred (S01-Yes) will be described. In this case, the control unit 110 determines in which direction panning is occurring (S02). This direction can be determined from the motion vector information used when determining the occurrence of panning. That is, the direction of panning can be determined from the direction of the vector. When it is determined that panning has occurred on the left side (S02-left), the control unit 110 sets C (1 + nK) as the reference camera (S04). Then, the control unit 110 determines the reference frame for each frame imaged by the reference camera as the nearest reference frame on the right side (S07). In other words, the control unit 110 sets the reference frame for each frame imaged by the reference camera as the nearest reference camera installed in the direction opposite to the direction in which panning occurs.

  On the other hand, when the control unit 110 determines that panning has occurred on the right (S02-right), C (N-nK) is set as the reference camera (S05). Then, the control unit 110 determines the reference frame for each frame imaged by the reference camera as the nearest reference frame on the left side (S08).

[Expanded decoding device]
Next, a configuration example of the decompression decoding apparatus 2 will be described. The decompression decoding apparatus 2 includes a CPU (Central Processing Unit), a main storage device (RAM), an auxiliary storage device, and the like connected via a bus in terms of hardware. The auxiliary storage device is configured using a nonvolatile storage device. The non-volatile memory device referred to here is a so-called ROM (Read-Only Memory: EPROM (Erasable Programmable Read-Only Memory), EEPROM (Electrically Erasable Programmable Read-Only Memory, etc.). ), Hard disk, etc.

  FIG. 7 is a diagram illustrating an example of functional blocks of the decompression decoding apparatus 2. The decompression decoding apparatus 2 loads various programs (OS, applications, etc.) stored in the auxiliary storage device into the main storage device and is executed by the CPU, whereby the header analysis unit 201, the encoded moving image buffer 202, It functions as a device including a variable length decoding unit 203, an IDCT inverse quantization unit 204, an adder 205, a display moving image buffer 206, a frame buffer 207, a motion / disparity vector compensation unit 208, a control unit 209, and the like. The header analysis unit 201, the variable length decoding unit 203, the IDCT inverse quantization unit 204, the motion / disparity vector compensation unit 208, and the control unit 209 are realized by executing a program by the CPU. The header 201, the variable length decoding unit 203, the IDCT inverse quantization unit 204, the motion / disparity vector compensation unit 208, and the control unit 209 may be configured as a dedicated chip. Next, each functional unit included in the decompression decoding apparatus 2 will be described.

<Header analysis section>
The header analysis unit 201 extracts header information from the input moving image data (moving image data created by the compression encoding device 1). Specifically, the header information includes the identifier of the camera that captured each frame, its type (base camera or reference camera), the identifier of the camera that captured the reference frame, and the like. Also, the header analysis unit 201 receives from the control unit 209 the identifier of the camera that captured the frame to be decoded. Based on this identifier, the header analysis unit 201 determines whether or not each input frame is to be decoded, and passes only the frame that is determined to be decoded to the encoded moving image buffer 202. Specifically, the header analysis unit 201 determines a GOP corresponding to a camera to be decoded among GOPs including frames captured by each camera (decoding camera determination process). By executing this process, it is possible to omit the decoding process for unnecessary frames, and to increase the processing speed. At the same time, the amount of data to be buffered can be reduced. For example, the scale of the display moving image buffer 206 and the frame buffer 207 can be reduced. Further, the header analysis unit 201 notifies the control unit 209 of the frame identifier passed to the encoded moving image buffer 202.

  Hereinafter, the decoding camera determination process will be described. In the following description, it is assumed that an image captured by the reference camera is compression-coded using only a frame captured by the standard camera as a reference frame. However, it is not necessary to be limited in this way. FIG. 8 is a flowchart illustrating an operation example of the header analysis unit 201 in the decoding camera determination process. The processing shown in this flowchart is executed every time SEQH is detected. First, the header analysis unit 201 acquires an identifier indicating a camera from the control unit 209 (S11). Next, the header analysis unit 201 extracts and analyzes the SEQH from the input moving image data (S12). By this analysis, the header analysis unit 201 acquires an error prediction type of each camera, an identifier of a camera serving as a reference destination, and the like.

  Next, the header analysis unit 201 searches for GOPH. At this time, if the next SEQH is detected, that is, if there is no frame imaged by the next camera, this processing is terminated (S13-No). On the other hand, when GOPH is detected (S13-Yes), the header analysis unit 201 analyzes the GOPH and acquires the identifier of the camera. Then, the header analysis unit 201 determines whether or not the identifier included in GOPH matches the identifier passed from the control unit 209. If they match (S14-Yes), it is determined that the decoding for each frame included in this GOP should be executed (S19). And the process after S13 is performed again.

  On the other hand, when the identifiers do not match (S14-No), the header analysis unit 201 determines whether the camera corresponds to the camera reference destination of the identifier passed from the control unit 209. When it is not a reference destination (S15-No), the processing after S13 is executed. On the other hand, when it is a reference destination (S15-Yes), the header analysis unit 201 analyzes GOPH and analyzes the time when the parallax prediction is not performed (S16). If the parallax prediction is performed for the entire time (S17-Yes), it is determined that the decoding for each frame included in this GOP should be executed (S19). On the other hand, when the parallax prediction is not performed in part of the time (S17-No), the header analysis unit 201 determines the frame on which the parallax prediction is performed, and determines that only that frame should be decoded ( S18). And the process after S13 is performed again.

<Encoded video buffer>
The encoded video buffer 202 buffers each frame determined to be decoded by the header analysis unit 201. The encoded moving image buffer 202 outputs a frame in units of decoding processing in accordance with an instruction from the control unit 209. The decoding processing unit may be one frame or a plurality of frames like GOP.

<Variable length decoding unit>
The variable length decoding unit 203 performs variable length decoding on the quantized DCT coefficient that has been subjected to variable length encoding, and passes the result to the IDCT inverse quantization unit 204. The variable length decoding unit 203 also performs variable length decoding on the motion vector information and the disparity vector information, and passes them to the motion / disparity vector compensation unit 208.

<IDCT inverse quantization unit>
The IDCT inverse quantization unit 204 performs an IDCT operation and an inverse quantization operation. The IDCT inverse quantization unit 204 performs inverse quantization and inverse DCT operation on the quantized DCT coefficient to obtain an inverse DCT operation result.

<Adder>
The adder 205 adds the inverse DCT calculation result and the prediction result by motion compensation or parallax compensation to generate a decoded moving image.

<Display video buffer>
The display moving image buffer 206 buffers the decoded moving image data generated by the adder 205. At this time, the display moving image buffer 208 buffers the data of the decoded moving image corresponding to the camera designated to be displayed from the outside. Then, the display moving image buffer 206 sequentially outputs the buffered data.

<Frame buffer>
The frame buffer 207 buffers the decoded moving image data in the same manner as the display moving image buffer 206. Further, the frame buffer 207 also stores frames captured by other cameras that are necessary for decoding the processing target frame, regardless of designation from the outside.

<Motion / parallax vector compensation unit>
The motion / disparity vector compensation unit 208 reads from the frame buffer 207 a frame used for motion prediction / disparity prediction necessary for decoding a frame to be processed in accordance with an instruction from the control unit 209. The motion / disparity vector compensation unit 208 acquires motion vector information / disparity vector information from the variable length decoding unit 203. Then, the prediction result is acquired, and the prediction result is passed to the adder 205.

<Control part>
The control unit 209 performs decoding control for input moving image data. In addition, the control unit 209 is connected so that instructions can be given to each functional unit. Further, an identifier indicating a camera that has captured a moving image to be output from the display moving image buffer 208 to the outside is input to the control unit 209. One or more identifiers may be used. Then, the control unit 209 passes this input identifier to the header analysis unit 201.

[Action / Effect]
[Modification]
Each frame (I ′ frame, P ′ frame, B ′ frame) imaged by the reference camera is not limited to a frame imaged at the same time by the reference camera, but refers to a frame imaged by another reference camera. Predictive encoding may be performed, or prediction encoding may be performed with reference to frames captured at different times by the base camera and other reference cameras.

  Further, the cameras do not need to be arranged in a straight line, and may be arranged in any manner such as a wavy line, a circle, a cross, or a square.

  The present invention can obtain an effect by being used for an apparatus that encodes and decodes images taken from multiple viewpoints.

DESCRIPTION OF SYMBOLS 1 Compression encoding apparatus 101 Input frame buffer 102 Subtractor 103 DCT quantization part 104 IDCT inverse quantization part 105 Adder 106 Frame buffer 107 Motion / disparity vector compensation part 108 Variable length encoding part 109 Header addition part 110 Control part 2 Decompression Decoding apparatus 201 Header analysis unit 202 Encoded moving image buffer 203 Variable length decoding unit 204 IDCT inverse quantization unit 205 Adder 206 Display moving image buffer 207 Frame buffer 208 Motion / disparity vector compensation unit 209 Control unit

Claims (2)

A decompression decoding device that decompresses and decodes moving image data captured and compressed by a plurality of cameras,
When decompressing and decoding moving images from a specific camera viewpoint,
Based on the information indicating whether or not the motion prediction using the moving image data captured by the camera of the other viewpoint is performed on the moving image captured by the camera of each viewpoint, the specific camera to be decompressed and decoded A viewpoint that captures the viewpoint moving image used for motion prediction at the time of compression encoding is determined for the viewpoint moving image data, and the moving image data captured by the camera of the viewpoint is used as the moving image data used for the motion prediction. A judging means for judging;
Decompression decoding means for performing decompression decoding on the moving image data of the viewpoint determined by the determination means and the moving image data captured by the specific viewpoint camera to be decompressed and decoded;
A decompression decoding apparatus. A decompression decoding method for decompressing moving image data captured and compressed and encoded by a plurality of cameras,
When decompressing and decoding moving images from a specific camera viewpoint,
Based on the information indicating whether or not the motion prediction using the moving image data captured by the camera of the other viewpoint is performed on the moving image captured by the camera of each viewpoint, the specific camera to be decompressed and decoded A viewpoint that captures the viewpoint moving image used for motion prediction at the time of compression encoding is determined for the viewpoint moving image data, and the moving image data captured by the camera of the viewpoint is used as the moving image data used for the motion prediction. A step of judging;
Performing decompression decoding on the viewpoint moving image data determined by the determination means and the moving image data captured by the specific viewpoint camera to be expanded and decoded;
A decompression decoding method.