JPH07327242A - Compressing and encoding device and decoding and reproducing device for stereoscopical animation picture - Google Patents

Abstract

PURPOSE:To reproduce and output natural picture data free from the degradation of the picture quality at the time of the occurrence of error or special reproducing. CONSTITUTION:Picture data of an object are obtained by an input means 1 of picture data for left eye and an input means 2 of picture data for right eye. These picture data are averaged by a reference picture data generating means 3 to obtain reference picture data, and it is compressed and encoded by an MPEG compression and encoding means 7. Each picture data has the motion vector detected by a motion vector detection means 5, and it is compared with reference picture data to perform motion compensation by a motion compensation means 6. Difference data is obtained by a difference means and is compressed and encoded. Two compressed and encoded data are subjected to timing adjustment by a timing control means 12 and are multiplexed by a data multiplexing means 11 and are recorded by a recording medium recording means 13.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a stereoscopic moving image compression encoding apparatus and stereoscopic moving image for obtaining a stereoscopic moving image by compression encoding a plurality of image data obtained by simultaneously photographing a subject from a plurality of directions. The present invention relates to an image decoding / reproducing device.

[0002]

2. Description of the Related Art Conventionally, various methods have been provided as methods for forming a stereoscopic moving image. The most common method is to alternately display images for the left eye and the right eye, which correspond to human eyes, in an image output device such as a television, and to use shutter glasses such as a liquid crystal shutter. . This method applies binocular parallax of the human eye. In order to obtain a stereoscopic moving image by such a method, one TV camera is placed at each of the left and right eye positions of a human, the respective image data are alternately recorded on the recording medium, and these display methods are performed. As for, the image data of the image output device such as a television is set to interlace,
Image data is alternately displayed on the left and right for each field or field group, and in synchronization with the switching, a stereoscopic moving image is obtained by alternately switching the left and right shutters set at the positions of human eyes.

However, these image data usually require twice as much data amount as two-dimensional image data, so that the recording capacity and the scale of the reproducing device tend to be large. In particular, when these image data are digitized, weaknesses such as increase in data amount and processing time have been pointed out.

Further, since these image data are obtained by photographing an object from a plurality of directions, the plurality of image data have correlations with each other and have similar data values. There are many things that are. Therefore, each image data is based on specific approximately similar image data, other image data is compared and referred to this reference image data, and the difference in data value is obtained to reduce the data amount, that is, Compression can be enabled. As such reference image data for comparison reference, a method of using any one of a plurality of image data captured at the same time has been conventionally used. For example, Japanese Patent Laid-Open No. 64-19892 discloses a method of comparing and referring to image data on one side and compressing and encoding the image data on the other side.

On the other hand, a digital video compression coding system has been remarkably developed in recent years, and in particular, a compression coding technique called MPEG (ISO 11172) system standardized by ISO has been adopted in various fields. There is. This is digital moving image data, one image unit from several frames to a dozen frames, fields, etc. is compression-encoded within the image unit, and other image data is other image data, that is,
This is a method of performing compression coding by difference and motion compensation obtained by comparing and referring to image data and the like in temporally front and back or different shooting directions.

[0006] As another example of the compression coding method using such a method, for example, H.264 is used. 261 or H.261. 22
There is a standard for videophone such as 1 and the like, and according to this, the data amount can be compression-encoded from the original fraction to several tenths, and in some cases to several hundredths, and , Relatively high image quality is obtained.

There are provided digital stereoscopic moving image compression techniques to which these techniques are applied. For example, JP-A-64-
5290, JP-A-64-5291, and JP-A-64-5292, compression coding is performed by using intra-image unit intra-image compression encoding and inter-image-unit compression encoding separately, and compression encoding is also performed. A method of decoding and reproducing the encoded data is disclosed.

Further, in Japanese Patent Laid-Open No. 4-361499, when one of the left and right eye image signals is subjected to intra-image unit compression encoding or inter-image unit compression encoding, it is generated when compression encoding is performed first. There is disclosed a method of compression-encoding using the other predicted image signal.

However, according to these techniques, for example, when an error occurs at the time of decoding, another image data that can be decoded is substituted and the reproduced image is obtained by reproducing the image data. As a result, the image becomes an unnatural stereoscopic moving image that gives the user a feeling of strangeness.

Such a thing will be described with reference to FIG.

FIG. 12 is an explanatory view for explaining a process when a decoding error occurs in a stereoscopic moving image reproducing apparatus which obtains a stereoscopic moving image by separately photographing right and left of a subject. In the figure, the subscripts of R are the image data for the right eye, and the subscripts of the L are the image data for the left eye.

A case will now be described in which the image data of 22R cannot be decoded due to a data reading error or the like. Only the image data 22L exists as the image data that can be decoded at the time of decoding the image data 22R. Therefore, the existing image data 22L is used as the image data to be decoded instead of the image data 22R, or nothing is decoded and reproduced. Then, as the reproduced image, appropriate image data is not displayed and output, so that the image becomes very unnatural.

Therefore, in general, the human visual characteristic in a stereoscopic image is a sensory expression in which the image data corresponding to the middle of the right and left image data is basically visually observed and a stereoscopic image is drawn in the head. Be expanded. Therefore, when an error as described above occurs, only two image data for the right eye or the left eye captured at the same time exist at the time when the error occurs, and the correlation of the image data Is the image data on the opposite side of the side where the error occurred, and therefore, as indicated by the arrow in the visual center of FIG.
The visual sensation is shifted to the right or left, and when the image data is output, the stereoscopic image is no longer collapsed, and the user feels uncomfortable. In order to avoid this, a method that can output image data closer to the visual center is required.

On the other hand, in the case where these image data are compression-encoded using intra-image unit compression and inter-image unit compression and special reproduction such as fast-forwarding is performed at the time of decoding and reproduction, intra-image unit compression-encoded image data A common method is to search for. This operation will be described with reference to FIG.

FIG. 13 is an explanatory view for explaining the decoding / playback sequence during fast-forward special playback in the conventional stereoscopic video decoding / playback apparatus.

In FIG. 13, shaded image data (33R, 36R, 39R, 42R, 32L, 35L,
38L, 41L, 44L) is image data that has been subjected to intra-image unit compression coding, and the other is image data that has been subjected to inter-image unit compression coding. In the MPEG system in which compression encoding is performed by intermixing intra-image compression encoding and inter-image compression encoding, inter-image compression encoding is performed by obtaining correlation with other image data. However, the compressed data itself that can be decoded is only the intra-image compression coded data. Therefore, at the time of special reproduction, the special reproduction is efficiently performed by extracting the compression-encoded data in the image unit from the sequence of all the compression-encoded image data and reproducing it.

In addition, the intra-image compression coded data extracted during the special reproduction may possibly be completely different image data among a plurality of image data. Therefore, as a fast-forward image in that case. Is not constant in the left-right direction, and may even cause the user discomfort.

A method of performing intra-image compression coding on only one of the left and right image data is also possible.
In this case, the inconvenience as described above does not occur, but since the image quality of the intra-image compression coded data is usually higher than that of the inter-image compression compression data, the image quality imbalance occurs. This causes a problem in the quality of the stereoscopic moving image.

[0019]

As described above, in these methods, an image in any one direction is used as a reference image at the time of compression encoding for forming a stereoscopic moving image by shooting from a plurality of directions. When compression encoding is performed by performing the difference and motion detection, an error in decoding occurs, or an unnatural image must be output during special reproduction, which causes deterioration in image quality. There was an inconvenience.

Therefore, the present invention provides a stereoscopic video compression encoding apparatus and a stereoscopic video decoding / reproduction apparatus capable of reproducing and outputting natural image data without image quality deterioration at the time of occurrence of an error or special reproduction. This is an issue.

[0021]

According to a first aspect of the present invention, there is provided a stereoscopic moving image compression coding apparatus, wherein the image data inputting means inputs image data obtained by photographing a subject from a plurality of directions, and the image. Reference image data generating means for comparing and referring to image data at the same time of each image data for each photographing direction obtained by the data inputting means, and one reference image data generating means, and a reference generated by the reference image data generating means. It is provided with a storage means for storing image data for a predetermined time, and a compression encoding means for obtaining compression encoded data by compression encoding the reference image data.

According to a second aspect of the present invention, there is provided a stereoscopic moving image compression coding apparatus, wherein the image data input means inputs image data obtained by photographing a subject from a plurality of directions, and the image data input means. Compare and refer to the image data at the same time of each image data for each shooting direction,
Change in movement between reference image data generating means for generating one reference image data, storage means for storing the reference image data generated by the reference image data generating means for a predetermined time, and image data compressed in the past A motion vector detecting means for detecting a motion vector, a motion compensating means for performing motion compensation by comparing each image data for each photographing direction input by the image data inputting means with the reference image data, the motion compensating means and the motion vector. An image compression encoding unit that obtains difference data by detecting a difference between each of the image data and the reference image data by the detection unit and compresses and encodes the difference data to obtain compression encoded data. It is a thing.

According to a third aspect of the present invention, a stereoscopic moving image compression encoding apparatus multiplexes the compression encoded data of the reference image data of claim 2 and the compression encoded data of the difference data to obtain a series of compression encoded data. A data multiplexing means and a recording medium recording means for recording the series of compression-coded data on a recording medium are provided.

According to a fourth aspect of the present invention, there is provided a stereoscopic video image decoding / reproducing apparatus in which recording medium reading means for reading out compression encoded data recorded in a recording medium and the compression encoded data is reference image data or difference data. Data determining means for determining whether the reference image data or the differential data is obtained by decoding the compression encoded data to obtain the reference image data or the differential data, and a storage means for storing the reference image data or the differential data for a predetermined time. And a motion compensation means for performing motion compensation by comparing and referring to the reference image data and the difference data, and an output control means for obtaining decoded image data for each shooting direction in the motion compensation means.

According to a fifth aspect of the present invention, there is provided a stereoscopic moving image decoding / reproducing apparatus in which recording medium reading means for reading out compression encoded data recorded in a recording medium and the compression encoded data are reference image data or difference data. Data decoding means for determining whether or not it is, decoding means for decoding the compression-coded data to obtain the reference image data or difference data, and an error occurs during decoding by the decoding means, making decoding impossible. In this case, an output control means for reproducing and outputting the reference image data instead of the reference image data is provided.

According to a sixth aspect of the present invention, there is provided a stereoscopic moving image decoding / reproducing apparatus in which a recording medium reading means for reading out compression encoded data recorded in a recording medium and the compression encoded data is reference image data or difference data. A data discriminating means for discriminating whether the reference image data is detected in high speed reproduction in a forward direction or a reverse direction, and decoding decoding means for decoding the detected reference image data; and the decoded reference image. And output control means for reproducing and outputting data to obtain a special reproduction image.

[0027]

According to the present invention, the image data for each shooting direction obtained by shooting the subject from a plurality of directions is obtained by each image data input means corresponding to the number of shooting directions. By comparing and referring to the obtained image data for each photographing direction, one reference image data is generated by the reference image data generation means, and is temporarily stored in the storage means. Then, after being stored for a predetermined time, it is input to the compression encoding means to obtain the compression encoded data of the reference image data.

In the second aspect, the image data for each shooting direction obtained by shooting the subject from a plurality of directions is obtained by each image data input means according to the number of shooting directions. Then, the image data for each shooting direction is input to the motion vector detection means, and a change in motion with respect to the reference image data that has been compression-coded in the past is detected as a motion vector.
Next, the motion compensation means calculates the difference between the reference image data and the image data for each shooting direction to obtain difference data.
The obtained motion vector and difference data are compression-encoded by the compression encoding means, and the image data for each shooting direction is compared with the reference image data to obtain the compression-encoded data of the difference data including the motion vector.

In the third aspect, the compression-encoded data of the difference data including the reference image data and the motion vector is
The data is multiplexed by the data multiplexing means to obtain a series of compression coded data. Then, the compression coded data is recorded on the recording medium by the recording medium recording means.

According to a fourth aspect of the present invention, the compression coded data of the reference image data recorded on the recording medium and the differential data including the motion vector is read by the recording medium reading means, and the compression coding is read by the data discriminating means. It is determined whether the data is reference image data or difference data. If the result of the determination is compression encoded data of the reference image data, it is input to the decoding means and decoded to obtain decoded image data. The compression-encoded data of the difference data is input to the motion compensating means and motion-compensated by analyzing the motion vector. Then, the difference data is decoded by the decoding means. The decoded difference data can be added to the reference image data stored in the storage unit to obtain the decoded image data for each shooting direction, which is reproduced and output by the output control unit.

In the present invention, the compression coded data is read by the recording medium reading means from the recording medium on which the compression coded data is recorded, and the data discrimination means discriminates whether the compressed data is the reference image data or the difference data. It Then, when an error or the like makes it impossible to decode the compressed coded data when it is decoded by the decoding means, the undecodable compressed data is replaced with the reference image data stored in the storage means. Then, it is reproduced and output by the output control means.

According to a sixth aspect of the present invention, when compression-encoded data recorded on a recording medium is subjected to special reproduction such as fast-forwarding, the reference image data discriminated by the data discriminating means is decoded by the decoding means and output. The control means reproduces and outputs.

[0033]

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram showing the configuration of a stereoscopic moving image compression encoding apparatus which is an embodiment of the present invention. 2 is a layout view showing the positional relationship between the image data input means and the subject, and FIG. 3 is an explanatory view for explaining the operation of the reference image data generation means of FIG. FIG. 4 shows the MPE of FIG.
It is a block block diagram which shows the structure of G compression encoding means.
FIG. 5 is an explanatory diagram showing a recording method of compression encoded data obtained by the stereoscopic moving image compression encoding apparatus of FIG. FIG. 6 is an explanatory diagram showing a data string of the compression-encoded data of the reference image data and the difference data multiplexed by the data multiplexing means of FIG.

In FIG. 1, reference numeral 1 is a left-eye image data input means for inputting left-eye image data of a video camera or the like, 2
Similarly, right eye image data input means for inputting right eye image data, 3 reference image data generating means for comparing and referencing left and right image data, and obtaining reference image data, 4 a frame memory for storing the reference image data 1 is a storage means, 5 is a motion vector detection means for detecting a motion vector, 6 is a motion compensation means, and 7 is an international standard MPEG (Mov
ing Picture Experts Group) ISO / IEC 11172-2 (M
PEG1), M compliant with ISO / IEC 13818-2 (MPEG2)
PEG compression / encoding means, 8 is DCT operation means for performing orthogonal transformation on input image data, 9 is quantization means for quantizing image data, 10 is DCT-ized or quantized data Is a variable-length coding means for allocating a short code to each and a long code to infrequent data. Further, 11 is a data multiplexing means for multiplexing the compressed data, and 13 is a recording medium recording means for recording the compression encoded data. Further, 12 is a timing control means for controlling the timing at which the compression encoded data is multiplexed, and 12a and 12b are buffer memories.

The subject 47 is photographed by the left-eye and right-eye image data input means 1 and 2 to obtain left and right image data, respectively. Such a photographing state is shown in FIG.

In FIG. 2, reference numeral 45 is a left-eye image data input means, 46 is a right-eye image data input means, and 47 is a subject. As the image data obtained by being photographed in the positional relationship with the subject 47, the left eye image data input means 45 obtains the image data 48 and the right eye image data input means 46 obtains the image data 49. Become.

The left and right image data 48, 49 thus obtained are sent to the reference image data generating means 3 in FIG. The reference image data generation unit 3 compares and references both the left and right image data captured at the same time and generates reference image data common to both image data. In addition,
The operation of the reference image data generating means 3 will be described later. Then, this reference image data will be used later in MPEG.
In order to absorb a processing delay and the like at the time of compression encoding by the compression encoding means 7, the frame memory 1 as the storage means 4 is temporarily stored for a predetermined time. Storage means 4
The reference image data output from is input to the MPEG compression encoding means 7 and compression encoded to obtain compression encoded data. The operation of the MPEG compression encoding means 7 will be described later. Then, the compressed coded data has a time difference when the left and right image data are compressed and coded by the timing control means 12,
In order to absorb this time difference, once the buffer memory 12
After being stored in a and 12b, they are sent to the data multiplexing means 11 and time division multiplexed.

On the other hand, the left and right image data are also input to the motion vector detecting means 5 to detect the left and right motion vectors, and the motion compensating means 6 predicts the motion of the left eye image and the right eye image. . Then, the motion compensating means 6 obtains the difference from the above-mentioned reference image data, and this difference value is sequentially processed by the DCT calculating means 8, the quantizing means 9 and the variable length encoding means 10 to obtain the reference image data and the difference data. The win control coded data is temporarily stored in the buffer memories 12a and 12b in order to absorb the time difference by the timing control means 12, and absorbs the difference in processing delay from the compression coded data of the reference image data described above. Then, the timing is controlled so as to correspond to the image data taken at the same time. Then, the timing-controlled compression coded data is time-division multiplexed by the data multiplexing means 11 and then recorded by the recording medium recording means 13 on a recording medium such as an optical disk or a magneto-optical disk.

In this embodiment, the image data photographed from the left and right positions, that is, the left eye and the right eye, are used. Accordingly, it is possible to expand easily by arbitrarily increasing the image data input means, the storage means 4, and the like.

Here, the operation of the reference image data generating means 3 will be described with reference to FIG.

In FIG. 3, (a) shows a process in which the reference image data is generated, and (b) shows an image of the image in each process. Further, (c) shows a method of forming the central image. In (a), 201 and 202 are a frame memory L for a left image and a frame memory R for a right image, respectively, 203 is a total motion vector detecting circuit, and 204 is a central image forming circuit.

Here, each image data obtained by the two image data input means 1 and 2 of FIG. 1 is temporarily stored in the frame memory L201 and the frame memory R202 of FIG. 3A. Then, the total motion vector detection circuit 203
In, the two image data are compared and the total motion vector V is detected as the motion vector shown in (b). This detects how much the center of gravity of the entire image is moving, for example, as compared with the image for the left eye and the image for the left eye. Next, the overall motion vector is calculated by dividing the entire image into a plurality of small blocks and obtaining the average value of each vector. Then, the central image forming circuit 204 averages the obtained total motion vectors V to form the central image shown in (b). Alternatively, in (c), for example, the common portion of the image R for the right eye and the image L for the left eye is extracted, and the lacking portion is filled with the corresponding retained portion of the image for the left eye as indicated by the hatched portion in the figure. Is also possible. The reference image data is generated by using the central image formed in this way.

Next, the above-mentioned MPEG compression encoding means 7
The operation of will be described with reference to FIG.

In FIG. 4, 110 is an image rearranging means for rearranging the reference image data in a time division manner, 112 is a scan conversion macroblock forming means for scanning-converting the reference image data into macroblocks, and 136 is a compression encoding form. The mode determination means 138 for determining whether or not is a motion detection means for detecting the motion amount of the reference image data before and after. As in FIG. 1, 134 is a motion compensating means, 116 is a DCT calculating means for performing a discrete cosine transform, 118 is a quantizing means, and 120 is a variable length coding means. And, 124 is a rate control means for controlling to keep the coding length constant. Further, in order to predict the difference from the next reference image data by using the previous reference image data and perform the compression encoding, a procedure reverse to the procedure of the compression encoding is provided, and the inverse quantization means 126, The inverse DCT calculation means 128 is used. Reference numeral 114 is a difference means for obtaining a difference between the image-unit compressed data and motion-compensated image data, and 130 is an adding means for adding the motion-compensated image data and the difference data of the decoded image-unit compressed image data. . 132 is an image memory means
It is divided into two regions 32a and 132b.

Therefore, the reference image data output from the storage means 4 in FIG.
At 10, the data are rearranged in the time series to obtain the time division data. The time-division data is subjected to scan conversion such as reducing the number of scan lines by the scan conversion macroblock forming means 112 to obtain macroblock data. Then, it is input to the mode determination means 136 and the compression coding form is determined.

In the case of performing intra-image compression encoding, intra-image compression encoding is performed via the path of the DCT calculating means 116, the quantizing means 118 and the variable length encoding means 120. The rate control unit 124 controls the amount of code generated after encoding to be a predetermined value for each encoding unit (one image, small area in an image, etc.), and the result is the quantization unit. Reached to 118. The data output from the quantizing means 118 is also sent to the inverse quantizing means 126. It is decoded into the original reference image data by the inverse DCT calculation means 128 and stored in the image memory means 132 (area of the image memory 132a). Then, the reference image data of the image memory 132a is input to the motion vector detecting means 5 and the motion compensating means 6 of FIG.

On the other hand, when performing compression coding between image units, first, the mode determining means 136 inputs the macroblock data to the motion compensating means 134, and the reference image data and macroblock data in the image memory 132a are input. The motion detecting means 138 compares the motion vectors to obtain a motion vector,
Similarly, it is input to the motion compensation means 134. Next, the motion compensating means 134 performs motion compensation on the macroblock data using the motion vector to obtain a difference value from the reference image data, and the difference value is stored in the image memory means 132 (area of the image memory 132b). Then, the difference means 114 subtracts this difference value from the macroblock data to obtain the inter-image unit difference data. The image-unit difference data is subjected to image-unit compression coding by the DCT calculating means 116, the quantizing means 118 and the variable length coding means 120 in the same manner as the reference image data. Similarly, the quantizing means 1
The data output from 18 is also sent to the inverse quantization means 126. The inverse DCT calculation means 128 decodes the original reference image data and stores it in the image memory 132a. Then, the reference image data in the image memory 132a is
It is input to the motion vector detecting means 5 and the motion compensating means 6 of FIG.

As described above, in this embodiment, the image data inputting means 1 and 2 for inputting the respective image data obtained by photographing the subject 47 from a plurality of directions and the image data inputting means 1 and 2 are used. Reference image data generating means 3 for generating one reference image data by comparing and referring to the obtained image data of each image data for each photographing direction.
From the storage means 4 for storing the reference image data generated by the reference image data generation means 3 for a predetermined time, and from the MPEG compression encoding means 7 for obtaining the compression encoded data by compression encoding the reference image data. It is provided with the following compression encoding means, which can be the embodiment of claim 1.

Further, the image data inputting means 1 and 2 for inputting the respective image data obtained by photographing the subject 47 from a plurality of directions, and the image data inputting means 1,
The reference image data generating means 3 for generating one reference image data by comparing and referring to the image data at the same time of the respective image data for each photographing direction obtained by 2 and the reference generated by the reference image data generating means 3. A storage unit 4 that stores image data for a predetermined time and a motion vector detection unit 5 that detects the amount of change in motion between image data compressed and coded in the past.
And a motion compensating means 6 for compensating motion by comparing each image data for each of the photographing directions input by the image data inputting means 1 and 2 with the reference image data, the motion compensating means 6 and the motion vector detecting means. 5, the image compression including MPEG compression encoding means 7 for detecting the difference between each of the image data and the reference image data to obtain the difference data, and compression encoding the difference data to obtain the compression encoded data. An encoding means is provided, which can be an embodiment of claim 2.

In this way, the image data inputting means 1 and 2 obtain the image data for each photographing direction obtained by photographing the subject 47 from a plurality of directions. The reference image data generating means 3 generates one reference image data by comparing and referring the obtained image data for each photographing direction, and temporarily stores the reference image data in the storage means 4 including the frame memory 1. After being stored for a predetermined time, it is input to the MPEG compression encoding means 7,
Obtain compression encoded data of the reference image data. At the same time, the motion vector detecting means 5 detects the image data for each shooting direction.
Is input to the CPU and the amount of change in motion with respect to the reference image data that has been compression-coded in the past is detected as a motion vector. Then, the motion compensation means 6 calculates the difference between the reference image data and the image data for each shooting direction to obtain the difference data.
The obtained motion vector and difference data can be compression-coded in the order of the DCT calculation means 8, the quantization means 9 and the variable length coding means 10.

As a method for compression-encoding the left and right image data, a method of performing motion compensation by referring to a reference image, obtaining a motion amount and a difference value thereof, and performing discrete cosine transform, that is, DCT, and compression-encoding. However, it is also possible to perform the compression encoding of the difference value without performing the DCT or the compression encoding without performing the motion compensation.

Next, a recording method of compressed coded data obtained by the stereoscopic moving image compression coding apparatus in the embodiment of the present invention shown in FIG. 1 will be described with reference to FIG.

In FIG. 5, (a) is image data for the right eye (1R, 2R, 3R, 4R, 5R), (b) is image data for the left eye (1L, 2L, 3L, 4L, 5L), ( Reference numeral c) is the reference image data (1I, 2B, 3B, 4P, 5B), and the numbers are the shooting order. Further, the frame memory 1 in (d) corresponds to the frame memory 1 in the storage means 4 in FIG. 1, and the frame memories A and B in (e) and (f) correspond to the image memory means 132 in FIG. The image data or the compression-encoded data stored in the table is shown. R'or L'indicates that the image data is compressed image data. Subscript I in the column of the reference image data shown in (c)
Represents the intra-picture compression coding data (Ipicture) in the MPEG system, the subscript P represents the forward inter-picture unit prediction data (Ppicture), and the subscript B represents the bidirectional inter-picture unit prediction data (Bpicture). Then, as described with reference to FIG. 3, the reference image data is generated by performing arithmetic averaging by adding the image data for the right eye and the image data for the left eye.

First, the reference image data generating means 3 shown in FIG. 1 arithmetically averages the right-eye image data 1R and the left-eye image data 1L to generate the reference image data 1I. The reference image data 1I is the MPEG image compression encoding means 7 of FIG.
The data multiplexing means 1 shown in FIG.
Sent to 1. At the same time, the right-eye image data 1R and the left-eye image data 1L are input to the motion vector detecting means 5 and a change in the amount of motion with respect to the compression-coded reference image data 1I in the past is detected as a motion vector. The motion compensation means 6 performs motion compensation to obtain a motion vector and difference data, and compression coding is performed via the paths of the DCT calculation means 8, the quantization means 9 and the variable length coding means 10. Then, the compressed coded data is temporarily stored in the buffer memories 12a and 12b by the timing control means 12 and then sent to the data multiplexing means 11, and the compressed coded data string shown in (g) by the recording medium recording means 13. As shown in FIG. 1, the compressed coded data 1R of the reference image data is followed by the compressed coded data 1R including the motion vector and the difference data.
And 1L.

Similarly, the reference image generation means 3 of FIG. 1 performs arithmetic averaging on the right-eye image data 2R and the left-eye image data 2L to generate the reference image data 2B. The reference image data 2B is sent to the MPEG image compression encoding means 7. On the other hand, image data 2R, 2L
Similarly to the above, when the motion vector is detected by the motion vector detecting means 5 and the motion vector and the difference data are calculated by the motion compensating means 6, it is stored in the frame memory 1 of the storage means 4 of FIG. 1 as shown in FIG. To be Next, the image data 3R and 3L are processed in the same manner, and as shown in FIG. 5, the reference image data 3B is stored in the image memory means 132 of the MPEG image compression encoding means 7, and the image data 3R and 3L are stored.
The motion vector and the difference data are stored in the storage means 4 including the frame memory 1.

Then, further, reference image data 4P is generated from the right-eye image data 4R and the left-eye image data 4L. The standard image data 4P is subjected to forward image unit compression encoding by the MPEG image compression encoding unit 7, and the compression encoded image data is stored in the image memory unit 132 of FIG. 4 as a reference image in FIG. Is stored as. At the same time, the motion vector and difference data of the image data 4R for the right eye and the image data 4L for the left eye are calculated and sent to the data multiplexing means 11 together with the compression coded data of the reference image data 4P for data multiplexing. Then, the data is recorded by the recording medium recording means 13.

After that, the reference image data 2B and 3B stored in the frame memory A are compression-coded between the bidirectional image units by the MPEG image compression coding means 7.
The compression-encoded reference image data 2B is data-multiplexing means 11 together with the compression-encoded data of the right-eye image data 2R and the left-eye image data 2L that have been compression-encoded and stored in the frame memory 1 before that. And data is multiplexed. Similarly, the compression coded data of the reference image data 3B, the right-eye image data 3R, and the left-eye image data 3L is sent to the data multiplexing means 11 and time division multiplexing is performed.

Thereafter, similar processing is performed, and the compression coded data shown in FIG. 5G is recorded by the recording medium recording means 13 in FIG.

Usually, in the MPEG moving image compression coding apparatus, in the case of intra-image unit compression or forward image unit compression, the image data is referred to when performing image unit compression of another image. The image data that has been subjected to this decoding process will be stored in the image memory means 132. However, in the case of bidirectional image unit compression, MPE
In the moving image compression by G, it is not referred to in the encoding of other image data, and it is not necessary to have the decoded image inside the encoder. However, in the stereoscopic moving image encoding apparatus of the present invention, even with regard to the image data for performing the compression between the bidirectional image units, since the decoded image is used for the left and right image encoding, the image data is also forward image unit. As in the case of the inter-compression coding, the decoding is performed inside the coding device, stored in the image memory means 132, and this image is used for left and right coding. For this reason, the image memory means 132 is required to have only the data for two images in the normal MPEG moving image coding apparatus, but at least for three images (compressed coded image within image unit, forward image unit It is necessary to have a memory for a compression coded image and a single bidirectional image unit compression coded image currently used for coding left and right data.

Next, the data multiplexing means 11 will be described with reference to FIG.

In FIG. 6, (h) shows the compression coded data of the reference image data, (i) the difference data for the left eye, and (j) the compression coded data of the difference data for the right eye, respectively. Show. (K) is a data string obtained as a result of time division multiplexing of these data by the data multiplexing means 11. Therefore, as shown in FIG.
A series of compression-encoded data that is regularly arranged in the order of (i) and (j) is obtained.

As described above, the compression coded data of the difference data including the reference image data and the motion vector is data-multiplexed by the data multiplexing means 11 to obtain a series of compression coded data. Then, the recording medium recording means 13 can be configured to record the compression-encoded data on the recording medium, and this can be the embodiment of claim 3.

Various methods other than the method of generating the reference image data by the arithmetic mean as in the present embodiment are possible. For example, when the photographing positions are two or more channels, it is possible to generate the reference image data by weighting the image data at the photographing position closest to the center direction where the subject 47 is photographed. Further, depending on the image capturing form of the object, for example, when the image is captured using an ultra wide-angle lens, a method of performing data conversion such as primary conversion of image data is also possible. In particular, when the average of a plurality of image data is taken as the reference image data, the image data of the target position,
The difference value with respect to the reference image data, on the other hand, is the same as the absolute value with the sign inverted, and the compression efficiency can be improved.

Next, a decoding / reproducing apparatus for decoding the compression coded data recorded on the recording medium according to the difference from the above-mentioned reference image data will be described with reference to FIG.

FIG. 7 is a block diagram showing the structure of a stereoscopic moving image decoding / reproducing apparatus for decoding the compression coded data recorded by the stereoscopic moving image compression coding apparatus shown in FIG. 1. FIG. 8 is an MPEG image shown in FIG. FIG. 9 is a block diagram showing the configuration of the decoding means, and FIG. 9 is an explanatory diagram showing a method of reproducing decoded image data obtained by the stereoscopic moving image decoding / reproducing apparatus in the embodiment of the present invention.

In FIG. 7, reference numeral 14 is a recording medium reading means for reading the compression coded data from the recording medium, and 15 is the compression coded data of the standard image data or the difference data. Data discriminating means for discriminating whether the data is data, 16 is an MPEG image decoding means for decoding the compression coded data of the reference image data, and 17 is an inverse DCT operation for decoding the left eye and right eye image compression coded data. Means, 18 is inverse quantization means, 19 is variable length decoding means, 20 is motion compensation means for analyzing the motion vector contained in the compression coded data, and 24 is the decoded right eye and left eye images. A storage unit composed of a frame memory 2 for storing data, and 22 is an output control unit for controlling the output of the reference image data, the image data for the right eye and the image data for the left eye to obtain a reproduced image. A. Reference numeral 23 is a data addition means.

Next, the operation of this embodiment will be described.

First, after the compression coded data read by the recording medium reading means 14 is sent to the variable length decoding means 19, the output data thereof is sent to the data discriminating means 15 and the compression code of the reference image data is sent. It is determined whether the data is encoded data or compression encoded data of difference data between the right-eye image data or the left-eye image data and the reference image data.

When the output rate from the recording medium and the decoding rate of the entire image are not the same in a short time (for example, when the decoding rate is not constant depending on the image), the reading means 14 and the variable length are used. A buffer memory for timing adjustment may be configured between the decoding means 19 and the decoding means 19.

Therefore, in the case of the compression encoded data of the reference image data, the MPEG image decoding means 16
To the output control means 22 after being transferred to and decoded. At the same time, the MPEG image decoding means 16 also inputs the decoded image data to the motion compensation means 20.

On the other hand, in the case of compression-encoded data of the right-eye image data and the left-eye image data, the data discriminating means 15 extracts the compression-encoded data from the motion compensating means 20 and the inverse quantization means 1.
Enter in 8. Inverse quantization is performed by the inverse quantization means 18,
The frequency domain data of the obtained difference component is input to the inverse DCT calculation means 17. Then, the inverse DCT calculation means 17 performs frequency-space domain conversion to obtain the difference component of the image data. Then, the decoded difference data between the right-eye image and the left-eye image is added by the adding means 23 in the motion compensating means 20.
The decoded image data generated by the addition of the motion-compensated image data obtained by the analysis of the motion vector is stored in the frame memory 2 of the storage means 24 once, and the processing time is adjusted. After that, the output control means 2
2, the reproduced image is obtained by performing control to display and output either the right-eye image data, the left-eye image data, or the reference image data.

In this embodiment, the left and right image data are motion-compensated to obtain the difference value, and the discrete cosine transform (DCT) is performed on the difference value to compress and encode the image data. When the image data is compression-encoded without performing DCT or when the image data is compression-encoded without performing motion compensation, it is necessary to perform decoding that matches the compression-encoding method.

Here, the MPEG image decoding means 1 of FIG.
The operation of No. 6 will be described with reference to FIG.

In FIG. 8, reference numeral 44 is a switch for switching the output depending on whether the image data to be decoded is intra-image unit compressed data or inter-image unit compressed data. Then, as in FIG. 7 or 4, 39 is an inverse DCT.
Arithmetic means, 40 dequantization means, 41 motion compensation means,
Reference numeral 42 is a storage means including a frame memory C for a reference image for motion compensation, and 43 is an addition means for adding the motion-compensated image data and the decoded difference data. Reference numeral 45 is a data separation means for separating motion vector data for motion compensation and data for difference values.

In FIG. 8, the following processing is performed on the compression coded data in image units. At the same time, it is input to the motion compensation means 41, but the switch 4
4 is connected to the upper side of the figure, the motion compensation means 41
Does not work. Then, the compression-encoded data output from the data separation means 45 is dequantized by the dequantization means 40, and the image data in the frequency domain is output. Then, this output is decoded by the inverse DCT calculation means 39 to decode the area data of the image. Further, the decoded reference image data is stored in the frame memory C of the storage means 42.

On the other hand, in the case of the inter-image compression coding, the switch 44 is connected to the lower side of the figure and the motion compensating means 41 is used.
Works. That is, the motion compensating means 41 obtains the difference value between the input compression-coded data and the reference image data stored in the storage means 42, which is compared and referenced and compression-coded between image units. Then, the inverse quantization means 40,
The compression coded data is decoded through the path of the inverse DCT calculation means 39 to obtain a difference value. In this way, the decoded difference data and the reference image data used by the motion compensating means 41 are added by the adding means 43 to obtain the reference image data compression-coded between image units.

Next, a method of reproducing image data in the stereoscopic moving image decoding / reproducing apparatus according to the embodiment of the present invention shown in FIG. 7 will be described with reference to FIG.

In FIG. 9, (m) shows the compression coded data read from the recording medium by the recording medium reading means 14 in FIG. 7, and (n) is stored in the frame memory C of the storage means 42 in FIG. Compression coded data,
(O) shows the decoded image data stored in the frame memory 2 of FIG. 7, respectively. Further, (p) and (q) indicate the right-eye image data and the left-eye image data reproduced by the output control unit 22. The subscripts of I, P and B are the same as in FIG. 5, where I is I picture, P is P picture,
B is B picture.

First, the compression coded data 1I is the MP of FIG.
The storage means 4 shown in FIG. 8 is decoded by the EG image decoding means 16.
2 is stored as a reference image in the frame memory C in FIG. Then, the compression coded data 1R and 1L are converted to the inverse DCT calculation means 17 of FIG.
The difference data is decoded through the inverse quantizing means 18 and the variable length decoding means 19. Then, the motion compensation is performed with reference to the reference image data 1I stored in the frame memory C, and the left and right image data 1R and 1L are decoded by being added to the difference data by the adding means 23 in FIG. 7, and the decoded image data To get After that, the decoded image data 1R and 1L
Is input to the output control means 22 of FIG.

Next, the compression coded data 4P is also decoded by the MPEG image decoding means 16 in the same manner as in the case of the reference image data 1I described above to obtain decoded image data. Then, the decoded image data is stored as shown in the frame memory C of FIG. The compression coded data 4R and 4L are also processed and decoded in the same manner as the above-mentioned image data 1R and 1L, but at this time, they are not sent to the output control means 22,
It is temporarily stored in the frame memory 2 of FIG.

Further, the compression coded data 2B is decoded by the MPEG image decoding means 16, and the decoded image data is stored in the frame memory C also in this case. Next, the difference data in which the compression coded data 2R and 2L are motion-compensated by the same process as the compression coded data 1R and 1L described above is decoded, and the decoded image data 2 stored in the frame memory C is decoded.
The motion compensation is performed with reference to B, the difference value is added, and the image data is decoded. Decoded image data 2R and 2L
Is input to the output control means 22. Thereafter, the compression coded data 3B, 3R and 3L are processed in exactly the same manner as 2B, 2R and 2L, and the decoded image data of 3R and 3L are input to the output control means 22. After such processing, the image data of 4R and 4L which have been decoded and accumulated in the frame memory 2 in the past are input to the output control means 22.

Thereafter, similar processing is performed, the decoded image data is input to the output control means 22 in the output reproduction order, and the processing shown in FIG.
The right-eye image data and the left-eye image data are reproduced as in (p) and (q).

As described above, the compression encoding data of the difference data including the reference image data and the motion vector recorded on the recording medium is read by the recording medium reading means 14 of FIG.
The data discriminating unit 15 discriminates whether the compression-encoded data read is the reference image data or the difference data. As a result of the discrimination, if it is the compression encoded data of the reference image data, it is input to the MPEG image decoding means 16 and decoded to obtain the decoded image data.

On the other hand, the compression-encoded data of the difference data is input to the motion compensating means 20 of FIG. 7 and motion-compensated by analyzing the motion vector. Further, the difference data is sequentially processed and decoded by the inverse DCT calculation means 17, the inverse quantization means 18, and the variable length decoding means 19. The decoded difference data is added to the reference image data stored in the frame memory 2 of the storage means 24 to obtain decoded image data for each shooting direction, and the output control means 22 reproduces and outputs the data. can do.

That is, the stereoscopic moving image decoding / reproducing apparatus of the present embodiment has a recording medium reading means 14 for reading the compression coded data recorded on the recording medium, and the compression coded data is reference image data, or Data discriminating means 15 for discriminating whether or not the data is differential data, and MP for decoding the compressed and encoded data to obtain the reference image data or the differential data
The motion compensation is performed by referring to the decoding means including the EG image decoding means 16, the storage means 24 for storing the reference image data or the difference data for a predetermined time, and the comparison reference of the reference image data and the difference data. Motion compensation means 20
And an output control means 22 for obtaining the decoded image data for each shooting direction in the motion compensation means 20, which can be the embodiment of claim 4.

In this embodiment, the reference image data is M
Although compression encoding and decoding are performed by the PEG method, in other cases, for example, when the reference image data is not compression encoded, the compression encoding and decoding of the reference image data according to each form is performed. May be.

Next, FIG. 10 shows the processing when an error occurs and decoding becomes impossible when decoding the compression coded data in the stereoscopic video decoding / reproducing apparatus according to the embodiment of the present invention.
Will be explained.

FIG. 10 is an explanatory view for explaining the processing when an error occurs in the stereoscopic video decoding / reproducing apparatus in the embodiment of the present invention.

In FIG. 10, image data 2 for the right eye is now taken.
When the 2R becomes undecodable due to a data reading error or the like, the right-eye image data 2 stored in the storage unit 24 in FIG.
Reproduction is possible by substituting the reference image data 22 for 2R. Therefore, as indicated by the arrow of the visual center of FIG. 10, the stereoscopic image that is smaller than the displacement of the visual center shown in FIG. 12 according to the related art and has a much higher image quality than the case of replacing with the left eye image data 22L. Can be obtained.

As described above, the compression coded data is read by the recording medium reading means 14 of FIG. 7 from the recording medium on which the compression coded data is recorded, and the data discrimination means 15 determines whether the compressed data is the reference image data or the difference data. Is determined.
Then, when an error or the like makes it impossible to decode the compressed and encoded data when they are decoded by the MPEG image decoding means 16, the reference image data stored in the storage means 24 is substituted and output. The control means 22 may be configured to reproduce and output.

That is, the stereoscopic moving image decoding / reproducing apparatus of the present embodiment has a recording medium reading means 14 for reading the compression coded data recorded on the recording medium, and the compression coded data is reference image data, or Data discriminating means 15 for discriminating whether or not the data is differential data, and MP for decoding the compressed and encoded data to obtain the reference image data or the differential data
Decoding means consisting of EG image decoding means 16 and output control for replacing with the reference image data and reproducing and outputting the reference image data when an error occurs during decoding by the decoding means and the decoding becomes impossible. And means 22 which can be an embodiment of claim 5.

Further, in the case of special reproduction such as fast-forwarding, generally, high-speed decoding and reproduction of compression-coded image data is required. Therefore, the processing for performing special reproduction in the embodiments of the stereoscopic video compression encoding apparatus and the stereoscopic video reproduction apparatus of the present invention will be described with reference to FIG.

FIG. 11 is an explanatory view showing the processing at the time of fast forward special reproduction in the stereoscopic moving image decoding / reproducing apparatus in the embodiment of the present invention.

In FIG. 11, reference image data (32, 35, 38, 41,
Reference numeral 44) is image data that has been subjected to intra-image compression coding, and the rest is reference image data that has been subjected to inter-image compression coding. Therefore, when performing fast forward playback,
Of the reference image data, only the reference image data, which is compression-coded in the image unit and shown by the diagonal lines, is decoded and reproduced. Therefore, far more efficient fast-forward special playback can be performed as compared with the case of performing the fast-forward special playback alternately shown in FIG. 13 according to the conventional technique.

In addition to the method described above, the inverse DCT operation requires a comparatively long time, so that the inverse DCT operation is required when decoding the left and right compression coded data.
A method of decoding and outputting only the difference data based on the motion vector without performing calculation is also possible.

As described above, when performing special reproduction such as fast-forwarding on the compression-encoded data recorded on the recording medium, as shown in FIG.
The reference image data discriminated by the data discriminating means 15 of M
A configuration in which the PEG image decoding means 16 decodes and the output control means 22 reproduces and outputs can be adopted.

That is, the stereoscopic moving image decoding / reproducing apparatus of the present embodiment has a recording medium reading means 14 for reading the compression coded data recorded on the recording medium, and the compression coded data is reference image data, or Decoding which comprises a data discriminating means 15 for discriminating whether or not it is difference data, and an MPEG image decoding means 16 for detecting the reference image data in high speed reproduction in the forward direction or the reverse direction and decoding the detected reference image data. It is provided with a reproducing means and an output control means 22 for reproducing and outputting the decoded reference image data to obtain a special reproduced image, which can be an embodiment of claim 6.

In this embodiment, the reference image data is M
It is compression-encoded by the PEG method, and during special reproduction such as fast-forwarding, only the intra-image compression encoded data of the reference image data is decoded and reproduced, but when the reference image data is not compression-encoded A method of decoding and reproducing image data every few images is also possible. Furthermore, the reference image data is MPE
When processing is performed by a compression encoding method different from the G method, special reproduction such as fast-forwarding is also possible by selecting compression encoded data to be decoded, performing decoding processing corresponding to the compression encoding processing, and reproducing. is there.

[0100]

As described above, in the three-dimensional moving image compression coding apparatus according to the first aspect, each image data for each shooting direction obtained by shooting the subject from a plurality of directions is obtained by the image data input means. The reference image data generating means generates reference image data by comparing and referring to all of the obtained plurality of image data captured at the same time. After that, the compressed coded data of the reference image data can be obtained by temporarily compressing and coding it by the compressing and coding means after being stored in the recording means. By using this reference image data, it is possible to reproduce and output natural image data without image quality deterioration when an error occurs or special reproduction.

In the three-dimensional moving image compression coding apparatus according to the second aspect of the present invention, each image data for each shooting direction is obtained by shooting the subject from a plurality of directions. One reference image data is obtained by comparing and referring to all the obtained plurality of image data captured at the same time. In the motion compensation means, the reference image data and each piece of image data are compared to obtain difference data, and each piece of image data is input to the motion vector means and compared with the reference image data compressed and encoded in the past. The motion change is detected as a motion vector, and the motion compensation of the difference data is performed. Then, the differential data including the motion vector is compression-encoded by the compression encoding means, and the compression-encoded data of the difference data can be obtained. As a result, it becomes possible to reproduce and output natural image data without image quality deterioration at the time of occurrence of an error or special reproduction.

In the three-dimensional moving image compression encoding apparatus of claim 3, the reference image data of claim 2 and the compression encoded data of the difference data are data-multiplexed by the data multiplexing means to obtain a series of compression encoded data. . Then, the compression coded data can be recorded on the recording medium by the recording medium recording means. Therefore, it is possible to reproduce and output natural image data without image quality deterioration at the time of occurrence of an error or special reproduction, and to reduce the data amount of compression encoded data without compressing and encoding all of the image data. Thus, the recording capacity of the recording medium can be used efficiently.

In the stereoscopic moving image decoding / reproducing apparatus of the fourth aspect, the compression encoding data of the difference data including the reference image data and the motion vector recorded on the recording medium is read by the recording medium reading means, and the data discriminating means is used. It is determined whether the read compression-coded data is the reference image data or the difference data. As a result of the determination, if the compression-encoded data is the compression-encoded data of the reference image data, it is input to the decoding means to be decoded and decoded. Image data can be obtained. The compression-encoded data of the difference data is input to the motion compensating means and motion-compensated by analyzing the motion vector. Further, the compression encoded data of the difference data is decoded by the decoding means, and the decoded difference data is added to the reference image data stored in the storage means to obtain the decoded image data for each shooting direction. You can A reproduction output can be obtained by inputting the decoded image data to the output control means. Therefore, it is possible to reproduce and output natural image data without image quality deterioration at the time of occurrence of an error or special reproduction, and since the amount of compressed coded data is small, it is possible to perform high-speed decoding and It is possible to obtain a high quality stereoscopic moving image.

In the stereoscopic moving image decoding / reproducing apparatus of the fifth aspect, the compression coded data is read from the recording medium on which the compression coded data is recorded by the recording medium reading means, and the compressed data is the reference image data by the data discriminating means. If it is impossible to decode due to an error or the like when these compression-coded data are decoded by the decoding means, the reference image data stored in the storage means is used. Instead, the output control means reproduces and outputs. This makes it possible to reproduce and output natural image data without image quality deterioration at the time of occurrence of an error or special reproduction, and to obtain a relatively high image quality even when undecoded compression coded data occurs. A high stereoscopic moving image can be obtained.

In the stereoscopic moving image decoding / reproducing apparatus of the sixth aspect, the data discriminating means is used when performing special reproduction such as fast-forwarding of the compression-encoded data recorded on the recording medium of the fourth or fifth aspect. The determined reference image data is decoded by the decoding means and reproduced and output by the output control means. This makes it possible to reproduce and output natural image data without image quality deterioration at the time of occurrence of an error or special reproduction, reduce the processing time, and efficiently perform special reproduction.

[Brief description of drawings]

FIG. 1 is a block configuration diagram showing a configuration of a stereoscopic moving image compression encoding apparatus which is an embodiment of the present invention.

FIG. 2 is a layout diagram showing a positional relationship between image data input means and a subject in the embodiment of FIG.

FIG. 3 is an explanatory diagram for explaining the operation of the reference image data generating means of the embodiment of FIG.

FIG. 4 is a block diagram showing a structure of an MPEG compression coding means of the embodiment shown in FIG.

5 is an explanatory diagram showing a recording method of compression encoded data obtained by the stereoscopic moving image compression encoding apparatus in the embodiment of FIG. 1. FIG.

6 is an explanatory diagram showing a data string of compression-encoded data of reference image data and difference data multiplexed by the data multiplexing means of FIG. 1;

7 is a block configuration diagram showing a configuration of a stereoscopic video decoding / reproducing device for decoding compression coded data recorded by the stereoscopic video compression encoding device of the embodiment of FIG. 1. FIG.

8 is a block diagram showing the structure of the MPEG image decoding means shown in FIG. 7.

FIG. 9 is an explanatory diagram showing a method of reproducing decoded image data obtained by the stereoscopic moving image decoding / reproducing apparatus in the embodiment of FIG. 7.

FIG. 10 is an explanatory diagram illustrating a process when an error occurs in the stereoscopic moving image decoding / reproducing apparatus in the embodiment of FIG. 7.

FIG. 11 is an explanatory diagram showing a process during fast-forward special reproduction in the stereoscopic moving image decoding / reproducing device in the embodiment of FIG. 7.

FIG. 12 is an explanatory diagram illustrating a process when a decoding error occurs in a stereoscopic moving image reproducing apparatus by separately capturing left and right of a subject.

[Fig. 13] Fig. 13 is an explanatory diagram for explaining a decoding / playback order during fast-forward special playback in the conventional stereoscopic video decoding / playback apparatus.

[Explanation of symbols]

1, 45 Left eye image data input means 2, 46 Right eye image data input means 3 Reference image data generation means 4, 24 Storage means 5 Motion vector detection means 6, 20, 41, 134 Motion compensation means 7 MPEG compression code Decoding means 8,116 DCT computing means 9,118 Quantization means 10,120 Variable length coding means 11 Data multiplexing means 114 Difference means 13 Recording medium recording means 14 Recording medium reading means 15 Data discriminating means 16 MPEG image decoding means 17, 39, 128 Inverse DCT calculation means 18, 40, 126 Inverse quantization means 19 Variable length decoding means 22 Output control means 23, 43, 130 Addition means 12 Timing control means 110 Image rearranging means 112 Scan conversion macroblocking Means 124 Rate Control Means 136 Mode Judging Means 138 Motion Detection Means

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁶ Identification code Office reference number FI technical display location H04N 11/04 B 9185-5C

Claims (6)

[Claims] 1. An image data input means for inputting image data obtained by photographing an object from a plurality of directions, and a plurality of image data obtained by the image data input means for each photographing direction at the same time. Reference image data generation means for comparing and referring to image data to generate one reference image data, storage means for storing the reference image data generated by the reference image data generation means for a predetermined time, and compression of the reference image data A stereoscopic moving image compression encoding apparatus, comprising: a compression encoding unit that obtains compression encoded data by encoding. 2. An image data input unit for inputting image data obtained by photographing an object from a plurality of directions, and an image data input unit obtained by the image data input unit for each image direction at the same time. Reference image data generation means for comparing and referring to image data to generate one reference image data, storage means for storing the reference image data generated by the reference image data generation means for a predetermined time, and compression-encoded in the past Motion vector detection means for detecting the amount of change in motion with image data, motion vector detection means for detecting motion with previously compressed and encoded image data, and for each of the shooting directions input by the image data input means. Motion compensation means for performing motion compensation by comparing each image data of the above with the reference image data; An image compression encoding unit that obtains difference data by detecting a difference between each image data and the reference image data, and compression-encodes the difference data to obtain compression encoded data. And a stereoscopic video compression encoding device. 3. Data multiplexing means for obtaining a series of compression coded data by multiplexing the compression coded data of the reference image data and the compression coded data of the difference data, and recording the series of compression coded data. 3. A recording medium recording means for recording on a medium.
The three-dimensional moving image compression encoding device according to [4]. 4. A recording medium reading means for reading compression coded data recorded on a recording medium, and a data discriminating means for discriminating whether the compression coded data is reference image data or difference data. Decoding means for decoding the compression-coded data to obtain the reference image data or difference data, storage means for storing the reference image data or the difference data for a predetermined time, the reference image data and the difference data A stereoscopic moving image decoding / reproducing apparatus, comprising: a motion compensating unit for performing motion compensation by comparing and comparing with each other; and an output control unit for obtaining decoded image data for each shooting direction in the motion compensating unit. 5. A recording medium reading means for reading the compression encoded data recorded on the recording medium, and a data discriminating means for discriminating whether the compression encoded data is reference image data or difference data. Decoding means for decoding the compression coded data to obtain the reference image data or difference data, and when the decoding by the decoding means causes an error and becomes undecodable, substitutes with the reference image data, An output control means for reproducing and outputting the reference image data, and a stereoscopic moving image decoding / reproducing apparatus. 6. A recording medium reading means for reading the compression coded data recorded on a recording medium, and a data judging means for judging whether the compression coded data is reference image data or difference data. Decoding / reproducing means for detecting the reference image data in high-speed reproduction in the forward direction or the reverse direction and decoding the detected reference image data, and reproducing and outputting the decoded reference image data to obtain a special reproduction image. A stereoscopic video image decoding / reproducing apparatus comprising: an output control unit.