JP2005245922A - Method, apparatus, and program for image data compression - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To efficiently compress image data carrying a plurality of sheets of tomogram and a frame image constituting an animation. <P>SOLUTION: The image data x1 to 8 carrying N (2≤N)-pieces of tomogram showing a subject sliced are encoded after converted by each block based on a prescribed conversion rule where reverse conversion exists and converted data may include the signal value of at least one low frequency component with N-pieces of image data about pixels whose positions are common to each other in the N-sheets of tomogram as one block. Then, encoded converted data f(y) 1 to 8 are gathered by each converted data obtained in the same process in the processes of conversion based on the conversion rule to obtain N-kinds of groups of converted data, and compression processing of the converted data f(y) 1 to 8 is carried out by each of the groups. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a method and apparatus for compressing image data carrying a plurality of tomographic images showing sliced subjects or a plurality of frame images constituting a moving image, and a program for causing a computer to execute the method. It is.

  2. Description of the Related Art Conventionally, CT (Computed Tomography) apparatuses, MRI (Magnetic Resonance Imaging) apparatuses, and the like that show a sliced portion of a subject such as a human body have been widely put into practical use. When obtaining a tomographic image of a predetermined part of a subject using this type of apparatus, in many cases, in order to know the three-dimensional state of the part, the depth direction (direction perpendicular to the image plane) A plurality of tomographic images are obtained by slicing at a plurality of locations. Naturally, the plurality of tomographic images obtained in this way are similar to each other.

  The tomographic image as described above is generally stored and stored for reexamination of a patient who is a subject, for example. In that case, hard copies of tomographic images may be stored and stored as they are, but recently, image data carrying these images is recorded and stored on a high-density recording medium such as a hard disk or optical disk, It is also widely stored. In order to record image data carrying a tomographic image on a recording medium using an image filing device in this way, normally, in order to record a larger number of image data, compression that suppresses redundancy of the image data is performed. Recorded as processed. In addition, such image data may be transferred between a plurality of departments via, for example, a transmission path provided in the hospital. In such a case, in order to shorten the time required for transfer, the image data Are transferred in compressed form.

  On the other hand, similarly to the above-described plurality of tomographic images, examples of the plurality of images that are similar to each other include the images of the respective frames constituting the moving image. Frame images constituting such a moving image are often recorded and transferred in the form of image data in the same manner as the tomographic image, but in that case, the image data is compressed and recorded and transferred. It is common.

Patent Document 1 shows an example of an apparatus for compressing image data carrying a tomographic image as described above.
Japanese Patent Laid-Open No. 10-127616

  By the way, the CT apparatus and the MRI apparatus described above have been improved so that the subject can be sliced thinner, and recently, an apparatus that can slice and show the subject every 1 mm is also available. Is provided. As described above, if the subject is sliced thinner, the number of tomographic images increases accordingly. Therefore, more efficient compression processing is required when recording and transferring image data carrying them. ing. As described above, more efficient compression processing is required for a plurality of frame images constituting the above-described moving image.

  Conventionally, when compressing a plurality of frame images constituting a moving image or image data carrying a plurality of tomographic images, compression processing is often performed for each of the images. It was. Paying attention to such a strong correlation between a plurality of images, several compression processes for suppressing redundancy between the images have been proposed. For example, the first tomographic image is encoded without using information between images, and the subsequent tomographic images are encoded with a difference value from the previous tomographic image (interframe prediction encoding). The method is known. However, while the conventional image filing apparatus is designed to handle one tomographic image as one file, in order to decode the tomographic image data encoded by the above-described method, a desired tomographic image is selected from the first tomographic image. Until the tomographic image is obtained, it is necessary to sequentially decode the encoded image data, and there is a problem that it takes a very long time to display the image. In addition, since the decoding apparatus is completely different from the conventional decoding apparatus, it is necessary to modify hardware and software in order to cope with such an encoding method. For this reason, when the conventional compression processing is applied, it has been difficult to reproduce one image representing them from the compressed data of a plurality of images.

  In the case of tomographic images, the slice thickness is set as thin as 1 mm at the time of photographing, the image data is taken into a workstation or a filing device, and at the time of film output, the slice thickness is set as thick as 8 mm for the purpose of reducing the number of films. Then, the pixels are reconstructed and printed. In addition, a thick slice image is used for normal film interpretation, and a thin slice image is used when creating a 3D image such as Volume Rendering. Thus, in the interpretation of tomographic images, an image representative of a plurality of tomographic images, for example, an image obtained by averaging a plurality of tomographic images may be required.

  On the other hand, in the case of a frame image constituting a moving image, if one image representing the plurality of images is reproduced, the one image is a so-called “frame drop-off” in which the time interval is shorter than the original moving image. "The video can be composed. As described above, even in the case of this moving image, it is difficult to reproduce the representative image with the conventional technique.

  In view of the above circumstances, the present invention can efficiently compress a plurality of tomographic images and image data carrying frame images constituting a moving image, and can easily generate a single image representing the plurality of images. Another object of the present invention is to provide an image data compression method that can be reproduced.

  Another object of the present invention is to provide an apparatus for performing the above-described method and a program for causing a computer to execute the method.

A first image data compression method according to the present invention includes:
A method of compressing image data carrying N (2 ≦ N) tomographic images shown by slicing a subject as described above,
The image data is inversely transformed with N image data relating to pixels having the same position in the N tomographic images as one block, and at least one low-frequency component is present in the converted data. The image data is converted for each block based on a predetermined conversion rule that includes the signal value of
Encoded conversion data is collected for each conversion data obtained in the same conversion process based on the conversion rule to obtain N conversion data groups,
For each of the groups, the conversion data is subjected to compression processing.

  More specifically, examples of the image data to be compressed by the first image data compression method according to the present invention include image data according to DICOM standard (Digital Imaging and Communication in Medicine).

A second image data compression method according to the present invention includes:
As described above, a method of compressing image data carrying N (2 ≦ N) frame images constituting a moving image,
The image data is inversely transformed with N image data relating to pixels having common positions in the N frame images as one block, and at least one low-frequency component is present in the transformed data. Is encoded after being converted for each block based on a predetermined conversion rule that includes the signal value of
Encoded conversion data is collected for each conversion data obtained in the same conversion process based on the conversion rule to obtain N conversion data groups,
For each of the groups, the conversion data is subjected to compression processing.

  In the first and second image data compression methods according to the present invention, N types of data after the compression processing are performed, and the remaining (N-1) types of data are converted into main one type of data. Is preferably recorded on the recording medium in the form of accompanying information. In this case, the main one-way data is preferably data composed of low-frequency components of the in-block image data before conversion.

  Moreover, as a preferable thing of said conversion process, orthogonal transformation etc. are mentioned, for example. On the other hand, as the above encoding, no conversion, scalar quantization, trellis code quantization, or the like can be preferably applied.

On the other hand, a first image data compression apparatus according to the present invention is for carrying out the first method,
In an apparatus for compressing image data carrying N (2 ≦ N) tomographic images showing a slice of a subject,
The image data is inversely transformed with N image data relating to pixels having the same position in the N tomographic images as one block, and at least one low-frequency component is present in the converted data. Means for converting the image data for each block based on a predetermined conversion rule that includes the signal value of
Means for collecting encoded conversion data for each conversion data obtained in the same process among the conversion processes based on the conversion rules to obtain N conversion data groups;
Each of the groups is provided with means for compressing the converted data.

A second image data compression apparatus according to the present invention is for carrying out the second method,
In an apparatus for compressing image data carrying N (2 ≦ N) frame images constituting a moving image,
The image data is inversely transformed with N image data relating to pixels having common positions in the N frame images as one block, and at least one low-frequency component is present in the transformed data. Means for converting each block based on a predetermined conversion rule that includes the signal value of
Means for collecting encoded conversion data for each conversion data obtained in the same process among the conversion processes based on the conversion rules to obtain N conversion data groups;
Each of the groups is provided with means for compressing the converted data.

  In the first and second image data compression apparatuses according to the present invention, N types of data after the compression processing are performed, and the remaining (N-1) types of data are converted into main one type of data. It is desirable to further include a means for recording the information on the recording medium in the form of accompanying information as accompanying information. In this case, the main one-way data is preferably data composed of low-frequency components of the in-block image data before conversion.

A first program according to the present invention is for causing a computer to execute the first method.
In a program for causing a computer to execute a method of compressing image data carrying N (2 ≦ N) tomographic images showing a slice of a subject,
The image data is inversely transformed with N image data relating to pixels having the same position in the N tomographic images as one block, and at least one low-frequency component is present in the converted data. A procedure for encoding image data after converting each block based on a predetermined conversion rule that includes the signal value of
A procedure of collecting encoded conversion data for each conversion data obtained in the same process among conversion processes based on the conversion rule to obtain N conversion data groups;
Each of the groups has a procedure for compressing the converted data.

A second program according to the present invention is for causing a computer to execute the second method.
In a program for causing a computer to execute a method of compressing image data carrying N (2 ≦ N) frame images constituting a moving image,
The image data is inversely transformed with N image data relating to pixels having common positions in the N frame images as one block, and at least one low-frequency component is present in the transformed data. A procedure of encoding after converting each block based on a predetermined conversion rule that includes the signal value of
A procedure of collecting encoded conversion data for each conversion data obtained in the same process among conversion processes based on the conversion rule to obtain N conversion data groups;
Each of the groups has a procedure for compressing the converted data.

  In the present invention, a case where orthogonal transform is applied as the transform processing and combined with high efficiency coding will be described as an example. First, an orthogonal transform encoding method for image data will be described. In this method, a numerical sequence consisting of sample values of digital two-dimensional image data is orthogonally transformed, and attention is paid to the fact that energy is concentrated on a specific component by this transformation. On the other hand, the low energy component is roughly encoded with a short code length to reduce the number of codes. As the orthogonal transform, Fourier transform, discrete cosine transform, Hadamard transform, Karhunen-Loeve transform, Haar transform, etc. are often used. The above method will be described in more detail by taking conversion as an example.

  First, as shown in FIG. 3, it is assumed that digital two-dimensional image data is divided into two blocks in a predetermined one-dimensional direction and divided into blocks each having a predetermined number of samples, and orthogonal transformation is performed for each block. When the two sample values x (0) and x (1) in this block are shown in an orthogonal coordinate system, they are highly correlated as described above, so that x (1) = x ( It will be distributed in the vicinity of the straight line 0). Therefore, this orthogonal coordinate system is transformed by 45 ° as shown in FIG. 4 to define a new y (0) -y (1) coordinate system.

In this coordinate system, y (0) indicates the low frequency component of the original image data before conversion, and y (0) is a value slightly larger than x (0) and x (1) (about 2 ^1/2). take ^twice) but, y (1) showing a high-frequency component of the one original image data will be only distributed in a very narrow range close to y (0) axis. Therefore, for example, if a code length of 7 bits is required for encoding x (0) and x (1), about 7 bits or 8 bits are required for y (0). For example, 1) can be encoded with a code length of about 4 bits. As a result, the code length per block is reduced, and image data compression is realized.

  As described above, the second-order orthogonal transform that constitutes one block for each of two image data has been described. However, as the order is increased, the tendency for energy to concentrate on a specific component becomes stronger, and the effect of reducing the number of bits is increased. Can do. In general, the above transformation can be performed by using an orthogonal function matrix. Ultimately, if the eigenfunction of the target image is selected as the orthogonal function matrix, the transformed image becomes its eigenvalue matrix, and the matrix pair The original image can be expressed with only the corner component.

  The transform data obtained by the above-described two-dimensional orthogonal transform is arranged in the order of the orthogonal function sequence (number crossing 0) used for the transform in each block. Since this sequence has correspondence with the spatial frequency, each conversion data is arranged in order of frequency. Therefore, a relatively long code length is assigned to the transformation data carrying the low frequency component (corresponding to the fact that a long code length is assigned to y (0) in the above-described one-dimensional quadrature orthogonal transformation), and the transformation data carrying the high frequency component. The code length per block is reduced by assigning or truncating a relatively short code length to.

  The plurality of tomographic images targeted by the first image data compression method according to the present invention are similar to each other because the subject is sliced at a thin interval as described above. That is, N pieces of image data related to pixels whose positions are common to each other in N pieces of tomographic images have high redundancy. Therefore, in this method, if these N pieces of image data are converted into one block and encoded after orthogonal transform for each block based on a predetermined orthogonal function, the code length per block for the reason described above. Is reduced, and data compression is realized.

  Further, in this method, the transform data encoded in such a high efficiency is collected for each transform data having the same sequence of the orthogonal functions to obtain N transform data groups. Since compression processing is performed on the data, very high compression efficiency can be obtained.

  Further, in the first image data compression method according to the present invention, the conversion data with the sequence of 0 (zero) is the DC component of the N image data in the one block (this is the average density of N pixels). Corresponding to the above). Therefore, the finally compressed data is subjected to decompression processing and decoding processing to obtain a conversion data group of this sequence 0, and by reproducing an image from the conversion data group, one sheet representing N tomographic images is obtained. A tomographic image can be easily obtained. As described above, this one representative tomographic image is an image obtained by slicing the subject relatively thickly.

  In this method, the final compression processing, that is, the compression processing performed on the converted data for each of the N groups, is based on a conventionally known lossless compression method or lossy compression method such as the JPEG standard. It is only necessary to apply compression processing or the like. Therefore, there is no particular difficulty in easily obtaining one tomographic image from the compressed data as described above.

  In addition, for the (N-1) group related to the sequence other than the sequence 0, the finally compressed data is subjected to the decompression process to obtain the converted data, and the converted data and the converted data of the sequence 0 are obtained. If the decoding process opposite to the encoding process and the inverse transformation process opposite to the orthogonal transformation are performed, all of the original N tomographic images can be reproduced.

  In this method, in particular, N types of data after being subjected to compression processing are recorded on the recording medium in a form in which the main (one) type of data is attached to the remaining (N-1) types of data as accompanying information. In this case, it is only necessary to manage one of N types of data, and management of image data is facilitated.

  Among them, particularly, when the main one type of data is data composed of low-frequency components of the intra-block image data before the conversion (sequential 0 conversion data in the case of the above-described orthogonal transform coding). The one kind of data indicates an image obtained by simply averaging N tomographic images for each pixel at a common position. That is, the image is a tomographic image sliced with a relatively large thickness obtained by multiplying the slice thickness of the original N number of tomographic images by N times. Since a tomographic image having a large slice thickness can be handled by reading it from the recording / reproducing apparatus, etc., interpretation work and the like are greatly facilitated. In such a case, when a tomographic image having a thin slice thickness is particularly necessary, the data carrying the thin slice thickness tomographic image is appropriately read out from the accompanying information and the tomographic image is reproduced.

  The effects described above can be obtained in the same way in the second image data compression method according to the present invention by replacing the “tomographic image” with the “frame image constituting the moving image”. Note that one representative image obtained in that case can be used to construct a so-called “frame dropping” moving image whose time interval is shorter than the original moving image.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

  FIG. 1 shows a schematic configuration of an apparatus for executing an image data compression method according to an embodiment of the present invention. For example, image data (original image data) x1, x2, x3, x4, x5, x6, x7, and x8 carrying a tomographic image output from the above-described CT apparatus, that is, an image showing a slice of a subject is stored in this apparatus. First, the data is passed through the preprocessing circuit 10 and subjected to preprocessing for improving data compression efficiency, such as smoothing for noise removal. The image data x1, x2, x3, x4, x5, x6, x7 and x8 are image data according to the DICOM standard as an example, and eight tomographic images F1, F2, F3 schematically shown in FIG. , F4, F5, F6, F7 and F8, respectively. In addition, these tomographic images F1 to F8 are shown by slicing the subject at intervals of 1 mm as an example.

The preprocessed image data x1 to 8 are passed through the orthogonal transformation circuit 11, and the pixels P _{(m, n)} 1, whose positions are common to each other in the eight tomographic images F1 to F8 in FIG. P _{(m, n)} 2, P _{(m, n)} 3, P _{(m, n)} 4, P _{(m, n)} 5, P _{(m, n)} 6, P _{(m, n)} 7 and P _{( m m, n)} Eight pieces of image data related to 8 are set as one block, and one-dimensional orthogonal transformation is performed for each block based on a predetermined orthogonal function. As this orthogonal transform, for example, a discrete cosine transform (DCT) is applied.

  The transformation data y1 to 8 obtained by the orthogonal transformation are arranged in the order of the function based on the orthogonal transformation in each block as shown in FIG. Since this sequence corresponds to the spatial frequency, the conversion data y1 to 8 are arranged in the order of the spatial frequency, that is, the order of the detail components of the image. In FIG. 2 (2), the closest conversion data y1 corresponds to the sequence 0 (zero), and the conversion data y1 indicates the average image density in the block.

  The converted data y1 to 8 arranged in this way are sent to the encoding circuit 12 shown in FIG. 1, where they are encoded. The encoding circuit 12 stores a predetermined allocation bit allocation table in, for example, a ROM, and encodes each conversion data y1 to 8 in the block with a code length (number of bits) according to the allocation table. . In the bit allocation table, a unique number of bits is assigned to each sequence, and the conversion data y1 to 8 are concentrated in low frequency components. By providing a relatively long code length, while providing a relatively short code length for high-frequency components with low energy, the required number of bits per block is reduced and image data compression is achieved.

  Here, in the encoding circuit 12, 0 (zero) bits may be given to the conversion data y having a sequence equal to or greater than a predetermined value and discarded. By doing so, all the converted data y showing a very high frequency component equal to or higher than a predetermined frequency, that is, data that has no significance in representing the original image is discarded, and the data compression effect is further enhanced. Such a technique is called so-called zone sampling.

  The encoded data f (y) 1 to 8 output from the encoding circuit 12 is subjected to a known compression process in the compression circuit 13. This compression processing is performed by the encoded data f (y) 1, f (y) 2, f (y) 3, f (y) 4, f (y) 5, f (y) corresponding to the above-mentioned sequence. 6, f (y) 7 and f (y) 8 are performed independently. As this compression processing, for example, compression processing according to the JPEG standard or the like is preferably used.

  The compressed encoded data f (y) ′ 1 to 8 is recorded in an image file made of a recording medium such as an optical disk or a magnetic disk in the recording / reproducing apparatus 14. Since the encoded data f (y) ′ 1 to 8 are subjected to significant compression on the original image data x1 to 8 by the orthogonal transform encoding and compression processing in the compression circuit 13, a recording medium such as an optical disk is used. A large amount of images can be recorded.

  When the tomographic image is reproduced, the encoded data f (y) ′ 1 to 8 are read from the recording medium of the recording / reproducing apparatus 14 and the decompressed circuit 15 encodes the encoded data f (y) before the compression. Converted to 1-8. The encoded data f (y) 1-8 are decoded into the converted data y1-8 by the decoding circuit 16. The transformed data y1 to 8 decoded in this way are sent to the inverse transformation circuit 17 and subjected to the inverse transformation for the one-dimensional orthogonal transformation, whereby the original image data x1 to 8 are restored. The original image data x1 to x8 are sent to the image reproducing device 18, and the eight tomographic images F1 to F8 carried by the data x1 to x8 are reproduced.

  The eight tomographic images F1 to F8 are shown by slicing the subject at 1 mm intervals as described above, but a tomographic image obtained by slicing the subject not so thin but thickly may be required. At that time, only the encoded data f (y) ′ 1 corresponding to the sequence 0 is read from the recording medium of the recording / reproducing apparatus 14, and the encoded data f (y) ′ 1 is read by the decompression circuit 15. It is converted into encoded data f (y) 1 before compression. Further, the encoded data f (y) 1 is decoded by the decoding circuit 16 into converted data y1.

The conversion data y1 obtained in this way is the DC component of the eight image data x1 to 8 (this is the pixels P _{(m, n)} 1, P _{( where the} positions are common to each other in the eight tomographic images F1 to F8 _{). m, n)} 2, P _{(m, n)} 3, P _{(m, n)} 4, P _{(m, n)} 5, P _{(m, n)} 6, P _{(m, n)} 7 and P _{(m, n) n)} corresponding to an average density of 8). Therefore, if the converted data y1 is sent to the image reproducing device 18 and the tomographic image carried by the data y1 is reproduced, the object is sliced every 8 mm.

  The encoded data f (y) 1 to 8 may be compressed by the compression circuit 13 by the same compression method. However, the encoded data f (y) 1 is applied, for example, by the above-mentioned JPEG standard compression processing. However, for the remaining encoded data f (y) 2 to 8, another compression method such as multi-stage run length encoding or bit plane encoding may be applied.

  Also, in a system that handles compressed encoded data f (y) '1 to 8, encoded data f (y)' 2 to 8 is additional data or private tag of encoded data f (y) '1. You may make it handle as information. Since the encoded data f (y) ′ 1 is shown by slicing the subject every 8 mm, as is apparent from the above description, in most cases where a tomographic image having a particularly thin slice thickness is not necessary, By handling only the encoded data f (y) ′ 1, data management, interpretation work and the like are facilitated. On the other hand, when a tomographic image having a thin slice thickness of 1 mm is necessary, the encoded data f (y) ′ 2 to 8 attached to the encoded data f (y) ′ 1 is also read out, and the slice thickness is 1 mm therefrom. The tomographic image may be reproduced.

The embodiment of the present invention using orthogonal transform coding has been described above. However, in the image data compression method of the present invention, image data transform other than orthogonal transform can be applied. An example of such another conversion is shown below. Here, the original image data carrying the tomographic image is assumed to be a ₁ , a ₂ , a ₃ , a ₄ , a _5, and reversible conversion is performed with 5 pixels as one block, and the converted image data is b ₁ , and _{b 2, b 3, b 4} , b 5. In the following formula,

Is the largest integer not exceeding x. Conversion

Do as. Decryption is

Do as. In this case, the conversion data b ₁ indicates the low frequency component, the conversion data b ₂ and b ₃ indicate the intermediate frequency component, and the conversion data b ₄ and b ₅ indicate the high frequency component. If the processing is applied, the same effect as that of the above embodiment can be obtained. Further, if the converted data b ₁ is compressed and the remaining converted data b ₂ is appended to the data obtained by compressing b ₅ as additional information, it is recorded on the recording medium, and this is also implemented. The same effect as that of the embodiment can be obtained.

  As described above, the embodiment applied when compressing image data carrying a tomographic image has been described. However, the image data compression method according to the present invention is applied when compressing the frame image constituting the above-described moving image. Is also possible. In this case, the conversion data y1 output from the decoding circuit 16 can be used to reproduce one frame image constituting a so-called “frame dropping” moving image whose time interval is shorter than that of the original moving image. .

  In the above-described embodiment, the image data x1 to 8 that respectively carry the eight tomographic images F1 to F8 are compressed. However, the image data compression method of the present invention can compress eight images. The present invention is not limited to this, and can be applied to all cases where two or more tomographic images or frame images constituting a moving image are compressed.

1 is a schematic configuration diagram of an apparatus for performing an image data compression method according to an embodiment of the present invention. Explanatory drawing explaining one process in the said image data compression method Explanatory drawing explaining the orthogonal transformation which concerns on this invention Explanatory drawing explaining the orthogonal transformation which concerns on this invention

Explanation of symbols

10 Pre-processing circuit
11 Orthogonal transformation circuit
12 Coding circuit
13 Compression circuit
14 Recording / playback circuit
15 Decompression circuit
16 Decoding circuit
17 Inverse conversion circuit
18 Image playback device

Claims (8)

A method of compressing image data carrying N (2 ≦ N) tomographic images showing a slice of a subject,
The image data is inversely transformed with N image data relating to pixels having the same position in the N tomographic images as one block, and at least one low-frequency component is present in the converted data. The image data is converted for each block based on a predetermined conversion rule that includes the signal value of
Encoded conversion data is collected for each conversion data obtained in the same conversion process based on the conversion rule to obtain N conversion data groups,
An image data compression method, wherein compression processing is performed on the converted data for each of the groups. A method of compressing image data carrying N (2 ≦ N) frame images constituting a moving image,
The image data is inversely transformed with N image data relating to pixels having common positions in the N frame images as one block, and at least one low-frequency component is present in the transformed data. Is encoded after being converted for each block based on a predetermined conversion rule that includes the signal value of
Encoded conversion data is collected for each conversion data obtained in the same conversion process based on the conversion rule to obtain N conversion data groups,
An image data compression method, wherein compression processing is performed on the converted data for each of the groups.   The N types of data after the compression processing are recorded on the recording medium in a form in which the main (one) type of data and the remaining (N-1) types of data are attached as accompanying information. The image data compression method according to claim 1 or 2. An apparatus for compressing image data carrying N (2 ≦ N) tomographic images showing a slice of a subject,
The image data is inversely transformed with N image data relating to pixels having the same position in the N tomographic images as one block, and at least one low-frequency component is present in the converted data. Means for converting the image data for each block based on a predetermined conversion rule that includes the signal value of
Means for collecting encoded conversion data for each conversion data obtained in the same process among the conversion processes based on the conversion rules to obtain N conversion data groups;
An image data compression apparatus comprising means for compressing the converted data for each of the groups. An apparatus for compressing image data carrying N (2 ≦ N) frame images constituting a moving image,
The image data is inversely transformed with N image data relating to pixels having common positions in the N frame images as one block, and at least one low-frequency component is present in the transformed data. Means for converting each block based on a predetermined conversion rule that includes the signal value of
Means for collecting encoded conversion data for each conversion data obtained in the same process among the conversion processes based on the conversion rules to obtain N conversion data groups;
An image data compression apparatus comprising means for compressing the converted data for each of the groups.   Means for recording the N data after the compression processing on the recording medium in a form in which the main (one) main data and the remaining (N-1) data are attached as accompanying information. 6. The image data compression apparatus according to claim 4 or 5, characterized in that: A program for causing a computer to execute a method of compressing image data carrying N (2 ≦ N) tomographic images showing a slice of a subject,
The image data is inversely transformed with N image data relating to pixels having the same position in the N tomographic images as one block, and at least one low-frequency component is present in the converted data. A procedure for encoding image data after converting each block based on a predetermined conversion rule that includes the signal value of
A procedure of collecting encoded conversion data for each conversion data obtained in the same process among conversion processes based on the conversion rule to obtain N conversion data groups;
And a program for performing compression processing on the converted data for each of the groups. A program for causing a computer to execute a method of compressing image data carrying N (2 ≦ N) frame images constituting a moving image,
The image data is inversely transformed with N image data relating to pixels having common positions in the N frame images as one block, and at least one low-frequency component is present in the transformed data. A procedure of encoding after converting each block based on a predetermined conversion rule that includes the signal value of
A procedure of collecting encoded conversion data for each conversion data obtained in the same process among conversion processes based on the conversion rule to obtain N conversion data groups;
And a program for performing compression processing on the converted data for each of the groups.

Images