JP2758159B2 - Image compression / decompression device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Object of the Invention] (Field of Industrial Application) The present invention relates to an image compression / expansion apparatus for compressing a captured original image and expanding the compressed image. (Prior art) When transmitting and receiving digital images through a communication line,
Alternatively, when storing in a storage device, or when reducing the load on a communication path or a storage device, the data amount is reduced by performing image compression processing on the original image. FIG. 5 shows a conventional image compression apparatus, which comprises an original image memory 1 for storing an original image, and a memory 1 for storing the original image.
And a compressor 2 for compressing the original image read from the original image memory 1 under the control of the address controller 4, and the compressed image is stored in storage means (for example, (Magnetic disk, optical disk) 3. The compressor 2 is, for example, as shown in FIG.
An input buffer 5 for temporarily storing data from the original image memory 1; a linear predictor 10 for performing linear prediction of data to be input next based on already input data; An adder 6 that performs an addition process with the predicted output of the predictor 10; a Huffman encoder 7 that performs a Huffman encoding process of the added output; and a byte format conversion process that performs a byte format conversion process of the Huffman encoded output. And an output buffer 9 for temporarily storing the converted output. Compressed image data is sent to the storage means 3 via the output buffer 9. The image compressed by the above image compression device is restored to the original image by going through the reverse process of the compression processing. This device is called an image decompression device, and basically has a decompressor 16 as shown in FIG. That is, the decompressor 16 comprises the input buffer 11 and the input buffer 11
A Huffman decoder 12 that performs a Huffman decoding process on the compressed image data captured via the Huffman decoding unit, an adder 14 that performs an addition process on the Huffman decoded output and the predicted output of the linear predictor 14, and A linear predictor 14 that performs a linear prediction process on data input next to the adder 13 based on the output,
An output buffer 15 for temporarily storing the addition output of the adder 13
And the output of the output buffer 15 is the image data restored to the original image. By the way, when the compressed image data is recorded on a storage device such as a magnetic disk or an optical disk, an error may occur when reading the data. For example, in the case of a magnetic disk or an optical disk, the unrecoverable error rate is 10 ¹² bits / error, and
An error occurs once in 25,000 × 10-bit images. Therefore, conventionally, as shown in FIG. 8, compression processing is performed for each line of the original image 17, and the processing result is written to the storage means 3 for each line. For example, as shown in FIG. When a data error occurs in a line (an error point is indicated by *), the propagation of the data error is suppressed to only the one line on the restored image 18. (Problems to be solved by the invention) However, even if the data error is 1 bit, since the compressed image data itself is coded,
On the restored image 18, not only the M-th line but also, for example, the M-th and subsequent lines as shown by diagonal lines in FIG. 10, adverse effects may occur, and important information may be lost over a wide range. . SUMMARY OF THE INVENTION Accordingly, an object of the present invention is to provide an image compression / expansion apparatus capable of obtaining a good restored image with little deterioration in image quality even when an error occurs in compressed image data. [Structure of the Invention] (Means for Solving the Problems) The present invention is directed to a case where a plurality of pixels which are not adjacent to each other on one line in a reading direction in an original image stored in a memory are defined as one pixel group. A dividing unit that divides all pixels in the original image into a plurality of pixel groups; a compressing unit that performs a compression process for each of the pixel groups; a storage unit that stores the compressed output; Restoring means for obtaining each restored image by decompressing the original image, means for restoring each restored image by the restoring means to the pixel array of the original image, and abnormal pixels on the restored image restored to the pixel array of the original image. And interpolating means for interpolating from the adjacent pixels. (Operation) All the pixels in the original image are divided into a plurality of pixel groups by the division means, compressed by the compression means for each pixel group, and stored in the storage means. For this reason, even if a data error occurs in any of the pixel groups in the compressed image data, it appears as an interspersed error on the image reconstructed by the reconstructing means, and the information of the specific area is completely lost. Never. Then, the error points scattered on the restored image can be easily interpolated by the interpolation means from the pixels adjacent to the error points. Therefore, even when an error occurs in the compressed image data, the image quality of the restored image is extremely small. (Examples) Hereinafter, the present invention will be described specifically with reference to examples. FIG. 1 is a block diagram showing one embodiment of the present invention.
As shown in the figure, the apparatus of this embodiment includes an image compression unit 20 for compressing an original image, a storage unit 21 for storing a compressed image, and an image expansion unit for expanding a compressed image read from the storage unit 21. 22. The image compression unit 20 includes an original image memory 23, an address controller 24,
It comprises a dividing means 25 and a compressing means 28. Original image memory 23
Stores an original image (uncompressed image) fetched from outside. The address control of the original image memory 23 is performed by an address controller 24. The dividing unit 25 divides all pixels in the original image into a plurality of pixel groups when a plurality of pixels that are not adjacent to each other on the original image are grouped into one pixel group, and is transmitted from the original image memory 23. It has a buffer memory 26 for storing data of one line of the original image, and a division control circuit 27 for enabling data transfer for each pixel group to the compression means 28 by address control of the buffer memory 26. The compression unit 28 performs a compression process for each pixel group, and the compressed image data generated by this process is stored in the storage unit 21.
It is stored in. As the storage means 21,
For example, a magnetic disk or an optical disk is applied. Further, the image decompression unit 22 includes a restoration unit 29, an interpolation unit 33, a restored image memory 34, and an address controller 35. The decompression means 29 is for decompressing the storage contents of the storage means 21 for each pixel group to obtain a decompressed image, and is a decompression device for decompressing the compressed image data sent from the storage means 21 for each pixel group.
30; a buffer memory 31 for storing the decompressed output; and a restoration control circuit 32 for rearranging data by controlling the address of the buffer memory 31. Interpolation means 33
Is to interpolate a data error portion on a restored image from adjacent pixels. Then, the formed restored image is sent to the outside via a restored image memory 34 arranged at the subsequent stage. The address control unit 35 controls the address of the restored image memory 34. Next, the operation of the above configuration will be described. <Image Compression> First, data of one line of the current image in the original image memory 23 is transferred to the buffer memory 26, and under the control of the division control circuit 7, the data of one line in the buffer 26 is converted into an odd number. The data is read out separately into data belonging to the address and data belonging to the even address. FIG. 2 (a), (b),
(C) shows the change of the address in this case.
When writing data to the buffer memory 26, addresses are specified in the order of 1, 2, 3, 4,..., N as shown in FIG. At the time of reading, as shown in FIG. 4B, odd address control for specifying addresses in the order of 1, 3, 5,..., N−1, and as shown in FIG. 4,6,…, N
, And the even-numbered address control in which the addresses are specified in this order. By such read address control, one line of data is divided into data belonging to odd addresses and data belonging to even addresses. If reduced, the pixel is divided into an odd-numbered pixel group and an even-numbered pixel group, which are composed of a plurality of non-adjacent pixels. Such a pixel group division process is performed on all lines for each line of the original image. And compression means
The compression processing is performed by 28 for each pixel group divided by the division means 25, and the processing results are sequentially sent to the storage means 21 and stored. FIG. 3 shows the relationship between the original image 17 and the compressed image data in the storage means 21. As shown in FIG. 3, the compressed image data obtained for each pixel group is stored in different areas in the storage means 21. Written. In the figure, "O" means that the data is an odd-numbered pixel group, and "E" means that it is an even-numbered pixel group. <Image Decompression> The compressed image data in the storage unit 21 is taken into the decompressor 30 for each pixel group, and after being decompressed, written into the buffer memory 31. In writing the decompressed data, the decompressed data for each pixel group is returned to the pixel array of the original image by the write address control by the restoration control circuit 32. That is, as described above, the decompression processing by the decompressor 30 is performed for each pixel group. As a result, the processed output is a pixel array for each odd address and each even address. By controlling the write address, the array of decompressed data on the buffer memory 31 is returned to the original pixel array before division. As a result, a restored image is obtained. FIG. 4 shows a restored image, and this pixel arrangement matches the original image. If an error occurs in any of the pixel groups in the compressed image data, the error always appears as an error scattered at every other pixel in the restored image 22, and the error completely eliminates information on the specific area. Not at all. Therefore, the abnormal pixel can be easily interpolated from the pixels (adjacent pixels) around the abnormal pixel by the interpolation means 33, and a good restored image with little image quality deterioration can be obtained. This restored image is sent to the outside via the restored image memory 34. Although one embodiment of the present invention has been described above, the present invention is not limited to the above embodiment, and various modifications can be made. For example, in the above-described embodiment, a description has been given of a case where the pixel is divided into pixel groups every other pixel, such as an odd-numbered pixel and an even-numbered pixel. Also,
The present invention can be applied to a case where a communication line is included in a transmission path of compressed image data. [Effects of the Invention] As described in detail above, according to the present invention, it is possible to provide an image compression / decompression apparatus capable of obtaining a good restored image with little image quality deterioration even when an error occurs in compressed image data. it can.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a block diagram of an apparatus according to an embodiment of the present invention, and FIGS. 2 (a), (b), (c), FIGS. 5, 6 and 7 are block diagrams of a conventional example, and FIGS. 8, 9 and 10 are explanatory diagrams of the operation. 21 storage means, 25 division means, 28 compression means, 29 decompression means, 33 interpolation means.

   ────────────────────────────────────────────────── ─── Continuation of front page    (56) References JP-A-63-73786 (JP, A)                 JP-A-61-167278 (JP, A)                 JP-A-59-50664 (JP, A)                 JP-A-61-157079 (JP, A)                 JP-A-60-128569 (JP, A)                 Showa 60-131061 (JP, U)

Claims (1)

(57) [Claims] When a plurality of pixels that are not adjacent to each other on one line in the reading direction in the original image stored in the memory are grouped into one pixel group, a dividing unit that divides all pixels in the original image into a plurality of pixel groups; Compression means for performing a compression process for each group; storage means for storing the compressed output; decompression means for decompressing the storage contents of the storage means for each pixel group to obtain respective reconstructed images; Means for returning the restored image to the pixel array of the original image, and interpolation means for interpolating abnormal pixels on the restored image returned to the pixel array of the original image from adjacent pixels. Stretching device.