JPH11196416A - Device and method for encoding, device and method for decoding and recording medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To efficiently perform encoding/combining processing of a picture at high speed. SOLUTION: The encoding device has the first and the second block generation parts 1 and 2 for dividing an input picture 100 into the first and the second block parts, a picture conversion/generation part 6 for applying a conversion/ generation to part of the second block, a proximity area search part 5 for deciding part of the converted block that is the most approximate to the selected first block, and an encoding/multiplexing part 8 for selecting the part of the second block corresponding to the block part decided and converted above, converting/multiplexing its position information and a conversion parameter for conversion and outputting them.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention generally relates to an encoding method and apparatus, a decoding method and apparatus, and a recording medium, and more particularly to a highly efficient image encoding or decoding system for efficient image transmission or storage. The present invention relates to an encoding method and apparatus, a decoding method and apparatus, and a recording medium that can be used.

[0002]

2. Description of the Related Art International Organization
zation for Standardization; ISO) is JPEG (Joint
Photographic Experts Group) has published a standardization system for conventional image compression. This system is called Discrete Cosine Transformation (DC)
Applying T) to the image to convert the image to DCT coefficients provides optimized coding or decoding of the image. This scheme works most efficiently when a relatively large number of bits are used to represent the encoded information. However, if the number of bits to represent the encoded information is less than a certain value, the inherent blockiness in such DCT transforms will be significant and the image quality will degrade as noticeably by the viewer. I do.

In response to these deficiencies in the JPEG and DCT procedures, a new Iterated Function Sy
stem; IFS) has been proposed and is gaining consent. This I
The FS technique focuses on the self-similarity between parts of an image and is based on fractal geometry.
IFS works under the assumption that the various parts of a particular image are similar, even if they are of different sizes (sizes), locations, perspectives or directions. IFS
Utilizes image redundancy to efficiently encode images without block distortions that may be generated in the JPEG scheme. Thus, the IFS has little dependence on the number of bits used to represent the encoded information, and the resolution during decoding uses a relatively small number of bits to represent the encoded information. They are not affected when they are sent.

The basic structure of IFS is described in Arnaud E. Jaquin, “Image Coding Based on a Fractal Theory of Iter.
atedImage Transformations) ”, IEEE Transactions on
Image Processing, Vol. 1, No. 1, pp. 18-30. Further, U.S. Patent Nos. 5,347,600 and 506, all issued to Barnsley et al.
No. 5447 and No. 4941193.
The encoding and decoding devices generally described in these references will now be described with reference to FIGS. 19 and 20 of the prior art.

Referring first to FIG. 19, the operation of an encoding device according to the prior art is shown. As shown in FIG. 19, the original image 300 is input to the block generation circuit 200, where it is divided into a plurality of blocks 301. All blocks 301 together completely cover the original image 300, but do not overlap each other. The original image 300 is also a reduced image generation circuit 202 that creates a reduced image 307 of reduced size by methods known in the prior art.
Sent to The reduced image is sent first and is stored in the reduced image storage circuit 204.

Each block 301 includes an approximate area search circuit 20
1, the approximate area search circuit determines whether there is a similar reduced image portion in the specific block 301 searched by searching the reduced image 307 stored in the reduced image storage circuit 204. As described above, this search involves searching for a portion of the reduced image 307 that has a different size, portion, perspective, or direction than the block 301 being searched. Approximate block position information 30 identifying a selected portion 305 in reduced image 307 according to a detected result indicating a successful search for the closest approximation.
6 is transmitted to the reduced image storage circuit 204. According to the result shown in this manner, the selected portion 305 of the reduced image 307 stored in the reduced image storage circuit 204 is extracted and transmitted to the rotation / inversion / level value conversion circuit 203.

The rotation / inversion / level value conversion circuit 203
The portion 305 of the reduced image 307 is converted to the approximate area search circuit 20
1 according to the conversion parameter 304 supplied from
Processing is performed by inversion / level value conversion. These conversion parameters 304 correspond to the selected part 3 of the reduced image 307.
The transformation applied to the selected portion 305 to transform 05 into block 301 is shown. These parameters are determined when a particular portion 305 of the reduced image 307 is found to correspond closest to the block 301 being searched. Following the conversion by the rotation / transformation / level value conversion circuit 203, the converted reduced image 3
03 is sent to the approximate area search circuit 201. As a result, the conversion parameter 304 and the approximate block position information 30
6 is output as the ISF code 302. Thus, a first image is input to the system and the output is a transformation parameter for transforming the first block of the first image into an approximated second block of the reduced image and the output of the encoded image. At least position information for determining the position of the second block is included.

Next, referring to FIG. 20, there is shown a decoding apparatus. The decoding apparatus includes, for example, an IFS recorded on a disk 400 including the conversion parameters output from the encoding apparatus shown in FIG. Code is IF
The data is input to the S code storage circuit 205 and stored. IFS
The code 302 is followed by the IFS code storage circuit 20.
5 is read for each block and sent to the IFS code reading circuit 206. IFS code reading circuit 206
Splits the code into approximate block position information 306 and transform parameters 304 as generated by the encoding device. The approximate block position information 306 is sent to the reduced image storage circuit 204 to reproduce the area of the reduced image specified by the approximate block position information 306. The reduced image portion 305 stored in the reduced image storage circuit 204 corresponding to the specified area is then transmitted to the rotation / inversion / level value conversion circuit 203 and according to the conversion parameter 304 supplied from the IFS code reading circuit 206. Is converted. The converted image 303 resulting from the conversion is rotated /
Sent from the inversion / level conversion circuit 203 and stored in the decoded image storage circuit 208. This procedure is performed for each block to which the IFS code has been given.

When all the IFS codes for all the blocks have been read, the IFS reading circuit 206 outputs a read end (READ OUT END) display signal to the copy control circuit 20.
Send to 7. Whether or not the information of all the blocks has been obtained is determined by the copy control circuit 207. Otherwise, a re-execute instruction 309 is sent to the IFS read circuit 206, the IFS code 302 is input again, and the procedure starts again. When the information of all the blocks has been read, the decoding continues by a recursive decoding procedure. The copy control circuit 207 counts the number of executed recursive decoding / copying operations. If this value does not reach a predetermined value, the copy control circuit 207 transfers the partially decoded image 313 to the information path 314. A reprocessing command signal 311 is sent to the switch 209 in order to send it to the reduced image generation circuit 202 via The reduced image generation circuit 202 rewrites the contents of the image stored in the reduced image storage circuit 204 and partially decodes the image 3
To enable the start of the next recursive decoding step at 15, a partially decoded reduced image 315 of the decoded image 315 is generated as in the encoding device. When a predetermined number of recursive decoding processes have been performed, and thus the copying operation has been performed a predetermined number of times, the decoded image 313 output from the decoded image storage circuit 208 is connected to the output port 316 so that the decoded image The output control signal is the reprocessing command signal 31
1 to the switch 209. The decoded image 313 is
After a predetermined number of recursive decodings, the control signal 31 consists of the whole of the above-mentioned decoded images.
2 is read out from the decoded image storage circuit 208.

[0010]

While this information encoding and decoding technique is somewhat satisfactory, it is at least obvious that the user will want to decode (encode) only a portion of the image. There are drawbacks. In the prior art described above, a range block (range
block) and a plurality of domain blocks located at arbitrary positions on the entire screen of the image are measured, and the domain block having the smallest error at first, and thus the closest block is selected. You. Thereafter, the transformation parameters and the block position information are determined, and these are given as codes. The codes are written in the bitstream in the order of these encoded blocks, such that the code for each range block makes up the entire image.

In order to further improve the efficiency of encoding and transmission, the size of the blocks of the domain blocks may vary, so that the words of the code are encoded in a variable length fashion to form a bit stream. Coding, ie, variable length Huffman coding or the like, is applied to the code or codes. However, in these cases, even if only a part of the image is decoded, it is necessary to read the entire bit stream from the header part to the appearance of the image in the predetermined area. In addition, since the domain blocks associated with the desired range block can be located anywhere in the image, the entire image needs to be decoded just to encode the desired portion of the image. Therefore, at the time of high-speed processing, there arises a problem that a considerable part of unnecessary information needs to be decoded in order to decode an image of a desired small part.

Accordingly, it would be advantageous to provide an improved encoding apparatus and method that overcomes the disadvantages of the prior art described above.

Accordingly, it is an object of the present invention to provide an improved encoding and decoding apparatus and method.

It is also an object of the present invention to provide an improved encoding and decoding apparatus and method which permits the decoding of only a portion of an encoded image as desired.

Still another object of the present invention is to provide an encoding and decoding apparatus in which a part of an image to be decoded is extracted from a bit stream containing encoded information of the entire screen without reading each part of the information of the bit stream. As well as a method.

The object of the present invention has been desired,
SUMMARY OF THE INVENTION It is an object of the present invention to provide a recording medium including all information for setting up an improved encoding and decoding method that allows only a part of an encoded image to be decoded.

It is another object of the present invention to provide an encoding and decoding method in which a portion of an image to be decoded is extracted from a bit stream including encoded information of a full screen without reading each part of the information of the bit stream. Is to provide a recording medium including all the information for setting.

[0018]

According to an aspect of the present invention, there is provided an encoding method for encoding an image, comprising the steps of dividing an input image into a plurality of first block portions. And a step of dividing the input image into a plurality of second block portions, and performing a conversion process on each of the second block portions to generate a plurality of converted block portions. And converting the most similar to the portion of the selected first block from among the plurality of converted blocks located in the same predetermined area as the selected first block of the input image. Determining a portion of the determined block, selecting a portion of the second block corresponding to the determined portion of the converted block, and determining a position of the selected portion of the second block. Bro showing The transformation parameters representative of said conversion processing portion of the second blocks click position information and the selection and a step of outputting a code.

According to the encoding method of the present invention, in the encoding method for encoding an image, the input image is divided into a plurality of regions, and the iterative transform encoding is performed on each of the plurality of regions in parallel. Performing the iterative transformation in parallel, dividing the input image into a plurality of first block portions, and dividing the input image into a plurality of second block portions. Dividing the second block into a plurality of converted blocks by performing a conversion process on each of the second blocks, and generating a plurality of converted block portions in the same manner as the selected first block of the input image. Determining a portion of the converted block that is most similar to a portion of the selected first block from among the plurality of converted blocks located in the predetermined area; and determining the determined converted block. Corresponding to Selecting the portion of the second block, converting block position information indicating the position of the selected portion of the second block, and a conversion parameter representing the conversion process of the portion of the selected second block. It has a step of outputting as a code and a step of multiplexing a code generated by each iterative transform coding.

[0020] A decoding method according to the present invention is a decoding method for decoding coded data composed of a plurality of codes.
A step of receiving block position information indicating the position of the transformation source block and a transformation parameter obtained by iterative transformation encoding of the block using only a predetermined area of the image, and a transformation process based on the transformation parameter, And performing a conversion process recursively on a block indicated by the block position information until a predetermined condition is satisfied for each block to be decoded in a predetermined area.

[0021] A recording medium according to the present invention comprises a step of dividing an input image into a plurality of first block portions on a recording medium on which a plurality of codes are recorded, and a step of dividing the input image into a plurality of first blocks. Dividing into two block parts;
Performing a conversion process on each second block portion to generate a plurality of converted block portions; and providing a plurality of converted block portions located in the same predetermined area as the selected first block in the input image. Determining, from the transformed blocks, a portion of the transformed block that is most similar to a selected portion of the first block; and determining a portion of the transformed block corresponding to the determined portion of the transformed block. A step of selecting a part of the second block, block position information indicating the position of the selected second block, and a conversion parameter representing the conversion processing of the selected second block as a code. The code generated by the repetitive transform coding in each step of the output step and the output step is recorded.

An encoding apparatus according to the present invention is an encoding apparatus for encoding an image, wherein the input image is converted into a plurality of first images.
A first dividing unit that divides the input image into a plurality of second block portions; and performs a conversion process on each of the second block portions. A conversion unit for generating a plurality of converted block portions, and selecting from among the plurality of converted blocks located in the same predetermined area as the selected first block of the input image. A determining unit that determines a portion of the converted block that is most similar to the portion of the first block that has been determined, and a selecting unit that selects a portion of the second block that corresponds to the determined converted block portion And an output unit that outputs, as codes, block position information indicating a position of the selected second block portion and a conversion parameter indicating the conversion process of the selected second block portion. Than it is.

An encoding apparatus according to the present invention is an encoding apparatus for encoding an image, comprising: an area determining unit for dividing an input image into a plurality of areas; and an iterative transform encoding for each of the plurality of areas in parallel. And a plurality of encoding units for performing the repetitive transform encoding in parallel, wherein the first encoding unit divides the input image into a plurality of first block portions.
, A second dividing unit that divides the input image into a plurality of second block parts, and a plurality of converted block parts that perform a conversion process on each second block part. And a conversion unit that generates a first block among the plurality of converted blocks located in the same predetermined area as the selected first block of the input image. A determining unit for determining a similar converted block portion; a selecting unit for selecting the second block portion corresponding to the determined converted block portion; Block position information indicating the position of the block portion and the selected second
And a multiplexing unit that multiplexes a code generated by each repetitive transform coding, and an output unit that outputs, as a code, a conversion parameter representing the above-described conversion processing of the block part.

A decoding device according to the present invention is a decoding device for decoding encoded data composed of a plurality of codes.
A receiving unit that receives block position information indicating a position of a transformation source block and a transformation parameter obtained by iterative transformation encoding of a block using only a predetermined region of an image,
By the conversion process based on the conversion parameter, in the predetermined region of the image, the conversion process is recursively performed on the block indicated by the block position information until a predetermined condition is achieved for each block to be decoded. And a conversion unit.

An encoding device according to the present invention is an encoding device for encoding an image, wherein the input image is converted into a plurality of first images.
A first dividing unit that divides the input image into a plurality of second block portions; and performs a conversion process on each of the second block portions. Converting means for generating a plurality of converted block portions, and selecting from among the plurality of converted blocks located in the same predetermined area as the selected first block of the input image. Determining means for determining a portion of the converted block most similar to the portion of the first block determined, and selecting means for selecting a portion of the second block corresponding to the determined converted block portion And an output means for outputting, as codes, block position information indicating the position of the selected second block portion and a conversion parameter indicating the conversion process of the selected second block portion. And it has a door.

An encoding apparatus according to the present invention is an encoding apparatus for encoding an image, comprising: area determining means for dividing an input image into a plurality of areas; and iterative transform encoding for each of the plurality of areas in parallel. And a plurality of encoding means for performing the iterative transform encoding in parallel,
First dividing means for dividing the input image into a plurality of first block portions; and dividing the input image into a plurality of second block portions.
A second dividing means for dividing the block into
Conversion means for performing a conversion process on the block portion of the input image to generate a plurality of converted block portions; and a plurality of the conversion portions located in the same predetermined area as the selected first block of the input image. Means for determining, from among the determined blocks, a portion of the converted block most similar to the selected portion of the first block; and the second block corresponding to the determined portion of the converted block. Selecting means for selecting a portion, and block code information indicating the position of the selected second block portion and a conversion parameter representing the conversion process of the selected second block portion are output as codes. It has output means and multiplexing means for multiplexing codes generated by each iterative transform coding.

[0027] A decoding device according to the present invention is a decoding device for decoding encoded data composed of a plurality of codes.
A receiving unit that receives block parameter information indicating a position of a source block and a conversion parameter obtained by iterative conversion coding of a block using only a predetermined area of an image, and a conversion process based on the conversion parameter,
Conversion means for recursively executing a conversion process on a block indicated by the block position information until a predetermined condition is satisfied for each block to be decoded in the predetermined area of the image.

According to the recording medium of the present invention, each code includes block position information indicating a position of a source block and a conversion parameter obtained by repetitive conversion coding of a block using only a predetermined area of an image. A recording medium on which a recording signal including encoded data composed of a plurality of codes is recorded, and a recording medium that can be decoded by a decoding device, wherein the block position information included in the encoded data and the encoded data are The step of receiving the included conversion parameter, and the conversion process based on the conversion parameter, the predetermined area of the image is indicated by the block position information until a predetermined condition is achieved for each block to be decoded. The recording signal decoded in each step of recursively executing the conversion process on the block is recorded. It is those that have been.

A recording medium according to the present invention includes a step of dividing an input image into a plurality of first block portions on a recording medium on which an encoding program is recorded; Dividing each of the second block portions to generate a plurality of converted block portions; and performing a conversion process on each of the second block portions to generate a plurality of converted block portions. Determining a portion of the converted block that is most similar to the selected first block portion from among the plurality of converted blocks located in the same predetermined area as the block; Selecting the second block portion corresponding to the transformed block portion; and selecting block position information indicating the position of the selected second block portion and the selected second block portion. In which the coding program of each step of the process of outputting the transformation parameter representative of said conversion processing portion of the block as the code is recorded.

According to the recording medium of the present invention, each code includes block position information indicating a position of a source block and a conversion parameter obtained by repetitive conversion coding of a block using only a predetermined area of an image. Receiving, on a recording medium on which a decoding program for decoding encoded data composed of a plurality of codes is recorded, block position information included in the encoded data and a conversion parameter also included in the encoded data, A step of recursively executing the conversion process on the block indicated by the block position information until a predetermined condition is achieved for each block to be decoded in the predetermined region of the image by the conversion process based on the conversion parameter. Are recorded.

[0031]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Preferred embodiments of the present invention will be described below with reference to the drawings.

First, referring to FIG. 1, an encoding apparatus according to an embodiment of the present invention receives an input image 100 and stores it, and stores the input image 100 in a first input image 101 and a first input image 101. First generating a plurality of first block portions, each receiving as a second input image
Block generation unit 1 and a second block generation unit 2 that generates portions of a plurality of second blocks, and a first input image 101 or a second input image 102 from the image memory 3
Control unit 4 that generates a control signal 103 for controlling the reading of
And In a preferred embodiment, the portion of the second block is twice as large as the portion of the first block. The plurality of second blocks received from the second block generator 2
Image conversion / generation unit 6 for converting the block part 106 of
Is provided. These converted block parts 107 are the second blocks received from the second block generator 2.
Of the first block 104 received from the first block generator 1 in the converted block portion 107
Is sent to the approximate block searching unit 5 for searching for the one most similar to. The approximate block search unit 5 outputs a control signal 105 to the control unit 4 in order to prompt the system to output the next image 100 from the image memory unit 3. The region determining unit 7 receives the input image 101 and divides the input image into a plurality of regions based on an external input. The first to be searched
Depends on the position of the block, the region determination unit 7 outputs the control signal 109 to output only a part of the input image 100 as the input image 102 corresponding to the region of the image 100 including the first block to be searched. Is output to the image memory unit 107. Therefore, in all cases, for all block parts 104, it is guaranteed that the correspondingly converted second block part 107 which is most similar is in the same area of the image 101.

The header information generation unit 29 outputs the header information 15
In order to generate 2, the number or address of the first block is received from the first block generation unit 1,
The number or address of the second block is received from the second block generation unit 2, and the screen division information is received from the same external input that gave this information to the area determination unit. The transformation of the selected second block portion 106 resulting in the first block portion number or address 111, the second block portion number or address 108, and the corresponding first block portion 104 is shown. Accompanying conversion parameter 11
0, the header information 152 is sent to the encoding / multiplexing device 8 which gives the final encoded output 120 recorded on the recording medium 400 or transmitted through the communication network 500.

Next, with reference to FIG. 4, the basic theory of iterative transform coding characteristic of the present invention will be described. Iterative transform coding is performed for all range blocks (range
This is a technique for executing a reduced mapping that is repeatedly performed from a domain block (domain block) to a range block. The mapping includes any number of transformations, including rotating, horizontally or vertically moving, enlarging or reducing, or changing perspective or direction of the domain block.
When the domain block closest to the range block is found, i.e., generating data with the least difference, the range block is encoded into an IFS code according to the procedure described below. Only the transform parameters and position information of the domain block most similar to the corresponding range block need be coded. As shown in FIG. 4, the range block Rk corresponds to the portion 104 of the first block, and the domain block Dk corresponds to the portion 10 of the second block.
6 is supported. The size of the range block Rk is m × n, and that of the domain block is M × N.
It is. As also shown in FIG. 4, there are L × L M first range blocks. Setting the size of the range block and the domain block is important because the size of these blocks has a significant effect on the coding efficiency.

The block image conversion procedure performed by the image conversion / generation unit 6 is conversion from block Dk to block Rk. Therefore, the mapping function of block k is wk,
Assuming that the number of blocks of the second block image required to map the entire image is P, the image f is mapped by the mapping function w of the entire image as Expression (1). W (f) = w ₁ (f) ∪w ₂ (f) ∪... ∪w _P (f) (1) Accordingly, W is represented by the following equation (2). W = ∪ _{k = 1} ^P W _k (2) It does not matter which type of mapping function is used as long as the function is a reduction function. Generally, reduced images are used to ensure convergence.

In addition, affine transformation is used from the viewpoint of simplicity of processing. The mapping from the block Dk to the block Rk by the affine transformation can be expressed as the following equation (3).

[0037]

(Equation 1)

Transformations between two blocks, such as rotation, horizontal or vertical movement, reduction or enlargement, can be fully described by equation (3) and are well known in the art. As shown, each parameter indicates a different transformation. Thus, (x, y) is represents the coordinates of a point before the affine _{transformation, (a i x + b i} y + e i, c i x + d i y
+ F _i ) represents the coordinates of the transformed point, which is v
_i (x, y).

The image conversion / generation unit 6 (FIG. 1) includes circuits or processing modules for conversions such as rotation, horizontal or vertical movement, reduction or enlargement, and is therefore represented by equation (3). You can perform all the functions performed. As an example, as shown in FIG. 4, the left corner above the domain block Dk is mapped to the left corner above the range block Rk. The conversion of the gray tone value of the pixel of the block image is realized in the same manner as the affine transformation described above.

The second block portion 106 read from the image memory section 3 and the second block generation section 2
, One or more transform coefficients (a _i , b _i , c _i , c _i ) of equation (3) in order to change the transformation applied to the second block portion 106 by the image transformation / generation unit 6.
d _i , e _i , f _i ) can be varied in various ways.
The plurality of second blocks generated by the second block generator 2
Block portion 106 is transformed in various ways to obtain a plurality of transformed block portions 107.
For these multiple transform block portions 107,
The approximate block search unit 5 detects the portion of the transformed block closest to the portion of the first block image 104.

The operation of the encoding device of FIG. 1 will now be described with further reference to FIG. The digitized image 100 is
The image data is input to the image memory unit 3 and stored therein. Under the control of the control unit 4, the first input image 101 is supplied to the first block generation unit 1 and the area determination unit 7. Area determination unit 7
Receives the screen division information 112 from an external source, and then determines the area of the second input image 102 to be read from the image memory unit 3 to the second block generation unit 2. This region is such that the portion of the second input image 102 sent to the second block generation unit 2 corresponds to the same region as the input image 100 where the first block to be tested from the first input image 101 is located. Is determined. For example, if a block read from the first image 101 and tried is located in the area 1 in FIG.
Of the input image 102 includes all portions of region 1, but does not include any portions of the image from other additional regions. In this example, only the portion of the input image 100 corresponding to the area 1 is sent to the second block generator 2. Thus, in the prior art, a domain block could be located anywhere in the image of a particular range block, but in the present invention, the domain block must be located as a block attached to the same area as the image 100. Therefore, the final second transformed block portion used for encoding the corresponding first block portion 107 is located in the same area as the input image 100. For example, FIG.
As shown in (2), the area determination unit 7 outputs the screen division information 112
Divides the screen vertically and horizontally as shown in FIG. In this example, the screen division information indicates that the screen is vertically divided into upper and lower parts and horizontally into left and right parts. It can be divided in any desired combination.

In this example, the screen is divided into four areas, area 1 to area 4, each of which has a horizontal width def.
_X_range and width def_y_ in the longitude direction
range. Further in accordance with the present invention, in the first region, a domain block (second block) D
A range block (first block) Rk corresponding to k is selected. In the second area, the domain block (second
The range block (first block) Rm corresponding to the block Dm is selected. In the third area, a range block (first block) Rn corresponding to the domain block (second block) Dn is selected. Finally, in the fourth area, a range block (first block) corresponding to the domain block (second block) is selected. During the operation, a transformation is performed on each domain block (second block) corresponding to the range block (first block) of the same area by mapping. As described above, the domain block corresponding to each particular selected range block is located in the same region as the corresponding range block. Thus, according to the present invention, iterative function coding is performed independently on each region of the image. The transform / encoding operation (shown in FIG. 4) is performed independently on each defined area of the image.

The conversion / coding operation performed in each area will be described with reference to the flowchart of FIG.

In the first step S1, the image memory unit 3
The first image is divided (generated) from the image 100 stored in the
To generate a plurality of first block portions 104.
In the next step S2, the second block 106 located in the same defined area of the input image 100 as the first block image is the image 1 stored in the image memory unit 3.
The second block 107 is divided (generated) from 00 and converted according to the defined conversion procedure, and the converted second block portion 107 is generated. Then, the process proceeds to inquiry S3.

In the inquiry S3, the current transformed second block portion 107 in the same area as the input image 100 as the selected first block portion 104 is read in step S3.
A comparison is made as to whether it most closely resembles a selected one of the plurality of first block portions 104 generated in 1.
This comparison is made by taking the sum of the absolute values between the pixels at each corresponding current first and second block position. The comparison between block positions is also made by using the sum of squares of the error in another embodiment. An inquiry is made as to whether the calculated error between portions of the current block is less than a predetermined threshold. If the answer is yes, this block portion is stored as one of the most similar block portion candidates in step S4, and the operation proceeds to inquiry S5. If the answer is negative, operation proceeds to inquiry S5.

An inquiry S5 determines whether all of the second block portions 106 located in the same region as the selected first block portion 104 being searched have been processed. All the second block portions 106 of this area
If has not been processed, operation returns to step S2, where a portion 106 of the second block next to the selected area of the image 100 is generated and moved according to step S2. However, if all the second block portions 106 of the given area have been processed, the operation will have a second error with the first error from the temporarily accumulated second block portion 106 candidates. The process proceeds to Step S6 where a block portion is selected. The operation is then performed with the conversion parameter 11 from the image conversion / generation unit 6 along with the number or address 111 of the first block part from the first block generation unit.
The process proceeds to step S7 in which the number address from the second block generation unit corresponding to the second block portion 106 selected as 0 is sent to the encoding / multiplexing unit 8. The operation then proceeds to inquiry S8.

The inquiry S8 determines whether all the first block parts 104 of the screen have been processed, or at least whether all the first block parts of the area of the screen that the user wants to encode have been processed. To determine. If this inquiry is answered negatively, control is returned to step S1,
Thus, the next first block portion 104 is generated.
If the query is answered affirmatively, the full screen encoded procedure ends.

Thus, according to the present invention, as shown in FIG. 2, each region is independently IFS coded according to the procedure described above for FIG.

Referring again to FIG. 1, the header information generator 29 converts the number or address 111 of the first block portion 104, the number or address 108 of the second block portion 106, and an external input source. Screen division information 11
Receive 2 In accordance with the screen division information 112, the header information generation unit determines the size of the screen, the number and size of the divided screens according to the screen division information 112, the number or address 111 of the first block portion 104, and the second block. Part 104 of each first block and part 10 of the second block according to the number or address 108 of the part 106 of FIG.
The information indicating the size of the data block representing the number 6 is generated. The length of each code is important to follow the invention,
A fixed length coding scheme is used rather than a variable length data size. Fixed length encoding allows random, immediate access to the address 108 of any particular first block portion because the location of the portion is known. Thus, the transformation parameters 110 of the second block portion 108 and each first block portion 104 are always located at the same position in the data stream and can be easily accessed. When a variable length coding scheme is used, it is necessary to read the entire stream of data to reach a particular selected block. In accordance with the present invention, when only certain portions of an image are decoded, immediately the necessary blocks of information in the encoded data stream to decode the appropriate image region including the portion of the image to be displayed. It is possible to move to an address.

Upon receiving the header information from the header information generation unit 29, the encoding / multiplexing unit 8 determines the number or address 111 of the portion 104 of the first block and the number of the portion 106 of the selected second block. Alternatively, it receives the address 108 and the conversion parameters 110 that are sequentially associated with each of the first block portions 104, and receives the header information 152, the conversion parameters 110, and the selected second corresponding to the first block portions 104. The number or address 108 of the block portion 105 is multiplexed in the order of the number or address 111 of the first block portion 104. In this method, encoding / multiplexing is performed by the encoding / multiplexing unit 8 in the order of writing by fixed-length encoding, and an encoded bit stream 120 is an output therefrom. This output coded bit stream 120 is stored in a recording medium 4 such as a so-called CD-ROM.
00 or transmitted to the communication network 50.

Referring to FIGS. 5 and 6, an encoding apparatus and method constructed in accordance with the present invention will be described. Elements that are similar to the above example will be indicated by similar reference numbers.

An encoding device constructed according to the present invention comprises:
A first block generator 1 for generating a plurality of first block portions 104 from the received input image 100 and a second block generation for generating a plurality of second block portions 106 from the received input image 100 2 includes a second block image memory unit 14 that stores a plurality of second block portions 106 generated by the second block generation unit 2. The encoding device further includes a second block image 10 received from the second block image memory unit 14.
6 includes an image conversion / generation unit 6 for performing a predetermined conversion on the second block portion 122. Further, a search is made for a portion 107 of the converted second block image received from the image conversion / generation portion 6 that is most similar to the portion 104 of the selected first block received from the first block generation portion 1. An approximate block search unit 13 is included. The control unit 4 also
The control signal 133 of the control for sending the converted second block portion 122 to the image conversion / generation unit 6 and performing a search there.
Is given to generate An area determination unit 7 is also provided for determining the area of the image 100 where the second image block portion 106 is supplied to the image conversion / generation unit 6.
This information is the screen division information 11 received from the external input.
2 and is determined by the region determining unit 7.

The area determining section 7 determines the first block part 1
A control signal 118 for selecting a portion 106 of the second block located in the same area as the image 100 corresponding to the image signal 04 is generated. The header information generation unit 29 is also selected by the approximate block search unit 13 as being most similar to the first block portion 104, with the number or address of the first block portion 111 from the first block generation unit 1. It is provided to receive the number or address 108 of the second block portion 106 and the header information 152 sent to the encoding and multiplexing unit 8, and the generated header information is sent to the encoding and multiplexing unit 8. The coding and multiplexing unit 8 sends the block numbers or addresses of the first and second block portions 111 and 108 and the conversion parameter 110 from the image conversion / generation unit 6 to the selected second block portion 106. The header information and the index of the conversion of the selected second block portion to the first block portion 104 are received from the header information generation unit 29 by the conversion. The encoding and multiplexing unit 8
The final encoded output 120 is recorded on the recording medium 400 or transmitted to the communication network 500.

The operation of the encoding device constructed in accordance with the present invention will now be described with reference to the flowchart of FIG.
In the first step S11, all the second block portions 106 that make up the image 100 are generated prior to the generation of any of the first block portions 104 and stored in the second image memory unit 14. Is held. Then, in step S12, the first block portion 1 of the encoding target (the first block image of the region of the image 100 to be decoded)
04 is generated. Therefore, in steps S11 and S12, the digitized image 100 is input to the first image generation unit 1 and the second image generation unit 2, where the first block portion 104 and the plurality of second blocks Are generated respectively. The operation proceeds to step S13. From step S13, the encoding technique in this example is similar to the technique discussed above. In particular, in step S <b> 13, the area determination unit 7 determines an area of the image 100 where the first block portion 104 is located from predetermined areas into which the entire screen is divided according to the image division information 112. The second block portion 122 located in the same region as the image 100 as the region where the first block portion 104 is located is the second block portion 122.
Are read from the block image memory 14 of the
The conversion is performed by the generation unit 6. The operation proceeds to step S14.

In steps S14 to S16, the approximate block searching unit 13 compares the second block portion 107 converted in step S13 with the first block portion 104. This comparison is made to determine which second block portion 122 has the first error between it and the first block portion 104 to be encoded. Of all of the second block portions 122 of the candidate region of the image 100, the second block portion having the smallest error between the second block portion 122 and the first block portion 104 The number or address 108 of 122 and the conversion parameter 110 are selected. This selection process is performed (as shown in each of the above examples) and the error for the transformed second block portion 107 to be tested is compared to a threshold. If the error is less than the threshold, operation proceeds to step S15, where the second block portion 122 is stored as a candidate for the most similar second block portion. The conversion parameter 100 and the number or address 10 of the selected second block portion 122
8 is temporarily stored. The operation proceeds to inquiry S16.
From step S14, if the error is smaller than the threshold or the previous minimum error, the operation proceeds to inquiry S16. Inquiry S16
Determines whether all second block portions 122 of the predetermined area, including the first block portion 104, have been processed. All the second in the defined area of the image 100
If the block part 122 is not processed, a control signal 119 is generated and transmitted from the approximate block search unit 13 to the control unit 4, and the control returns to step S13, where the control signal 133 generated by the control unit 4 Accordingly, the second block image portion 122 is stored in the second block image memory unit.
Is read. When all the second image portions 122 in the same image 100 region as the first block image 104 have been processed, a control signal 119 is generated and transmitted from the approximate block search unit 13 to the control unit 4. The operation then proceeds to step S17.

In step S 17, in response to the control signal 133, the second selected block having the minimum error between the conversion parameter 110 and the data of the second block and the first block 104 is converted. The number or address 108 of the part 122 is the second temporarily stored number.
Is selected from all the candidates of the block part 106 of the block.
The operation then proceeds to step S18, where the selected conversion parameter 110 from the image conversion / generation unit 6, the selected number or address 108 of the selected second block part 122 and the number of the first block part are selected. Alternatively, the address 111 is sent to the coding and multiplexing unit 8. Operation proceeds to step S19, where an inquiry is made as to whether all the first block portions 104 of the image 100 have been processed. Otherwise, control returns to step S12 to generate the first block portion 104 to be added. When all the first block portions 104 have been generated, the entire screen has been coded, and this procedure ends.

Referring now to FIG. 7, an encoding device constructed in accordance with the present invention will be described. The encoding device shown in FIG. 7 is similar to the encoding device described above,
Extraction unit 11 for extracting a region from an image 100 in place of
5 and a local memory unit 116 for storing the area extracted by the area extracting unit 115. The rest of the elements are similar to those described in the previous example, and are therefore indicated by similar numbers. The operation of an encoding device constructed in accordance with the present invention will now be described. In this encoding device, the digitized image 100 is input to the image memory unit 3. Only a specific area 124 of the image 100 is read from the image memory unit 3 according to the image division information 112. Image memory unit 3
The region 124 of the image 100 read from is transmitted to the local memory 116 as the region image 125. According to the read control signal 103 received from the control unit 4, the first image 126 and the second image 127 are output from the local memory unit 116 to the first block generation unit 1 and the second block generation unit, respectively. is there. The processing of the first block portion 104 and the second block portion 106 generated by the first and second block generation units 1 and 2, respectively, is quite similar to the above-described processing. When all of the first block portions 104 included in the region image 125 are encoded, the other region images 12
5 is extracted from the image memory unit 3 in the same manner by the area extracting unit 115 according to the read signal 123 from the control unit 4. The additional portion is stored in the local memory 116, and the same operation as described above is performed.

Referring now to FIG. 8, an encoding device constructed in accordance with the present invention will be described. As shown in FIG. 8, the encoding device constructed in accordance with the present invention includes an image memory unit 3 for storing an input image 100 and a plurality of predetermined screens of the image 100 according to screen division information 112 received from an external source. A screen division unit 22 for dividing the image into regions, a first region repetition function encoding unit 23 that performs iterative variable encoding on a first predetermined region, and a second unit that performs an iterative function transformation on a second predetermined region. 2, an area repetition function encoding unit 24,
And an other-region repetition function encoder typified by a K-th region repetition function encoder 25 for performing iterative transform encoding on the K-th region. The elements included in each iterative function encoder are the same. The multiplexing unit 26 is provided for multiplexing the header information and the coded bit stream from each predetermined area.

The operation of the encoding device of FIG. 8 will now be described. In this encoding device, after the digitized image 100 is input and stored in the image memory 3, the image 100 is sent to the screen division unit 22, and the entire screen of the image 100 is processed according to the division information 112 received from the external input. , Divided into multiple screens. As an example, in FIG. 8, the screen is divided into K areas. Images 134 and 14 of divided areas
, 136 are sent to the first, second,..., K-th iterative function encoders 23, 24,. In each region, the iterative function transformation is performed independently. As an example, if K is 4, the input image is divided into four regions. The first iterative function encoding unit 23 performs region 1
, And a second iterative function encoder 24 performs an iterative function transform on the image in region 2 and so on, as shown in FIG. Transformation part 2
5 performs iterative function conversion on the image of the area 4. The coded bitstreams 137, 138, ..., 139 are the first,
The second,..., K-th area repetition function encoders 23, 2
, 25 are output. The image division information 112 is sent to the multiplexing unit 26 together with the encoded bit stream. The multiplexing unit 26 then generates header information, and outputs the header information and the encoded bit stream 137,
138 and 139 are multiplexed, and the multiplexed information is output as a bit stream 151.

FIG. 9 shows the internal structure of the iterative function encoder 23. Each of the other area repetition function encoder waves has a similar internal structure. In the region repetition function encoder shown in FIG. 9, the first region image is input to the local memory unit 27, and when prompted by the signal 143 from the approximate block search unit 5, the first image 140 and the second image 141. Is the local memory unit 2 according to the control signal 142 from the control unit 4.
7, the first block generator 1 and the second
Are output to the block generation unit 2. The rest of the process is similar to that described in the above example, with multiple first
, Generation of a plurality of second block portions 145, conversion of the second block portions 145 into the second block portions 147 converted by the image conversion / generation unit 2, approximation It includes the determination of the minimum error by the block search unit 5, the multiplexing by the encoding / multiplexing unit 8, and the output as an encoded data stream of information. This example is characterized in that high-speed encoding is performed by a plurality of encoding units in parallel with respect to encoding performed in succession to the preceding example.

Next, an embodiment of the encoding method according to the present invention will be described with reference to FIG. The flowchart described in FIG. 10 could be programmed in a general computer. In this program, in a first step S21, an input image is stored in a frame memory of a general computer. In the next step S22, the image accumulated in step S21 is divided to generate a plurality of first block portions. In the following step S23, the image accumulated in step S21 is similarly divided into a plurality of second images.
Generate the block part of. Then, the operation proceeds to step S24, in which all the second block portions belonging to the same region as the respective first block portions are determined. Then, in step S25, one first block accumulated and generated in step S22, and all the second block portions located in the same predetermined area as the first block are read. Then, the operation proceeds to step S26, and each second block read in step S25 is converted. In the subsequent step S27, which second
Is the closest to the portion of the selected first block, ie, which second block portion has the least error between it and the selected first block portion Is performed. This process is performed using the threshold processing in the example of the state j, and in step S28, the portion of the second block having the minimum error is selected.
The number or address and the conversion parameter corresponding to the selected second block portion and the number or address of the first block portion are generated and accumulated. The operation is
Then, the process proceeds to a query S29 for determining whether or not all the first block portions of the input image have been encoded. If all the first block parts have not been encoded, the operation returns to step S25, and the next first block part is read out and the operation that proceeds is executed. All first
If the block has been encoded, the operation proceeds to step S30, and as shown in FIG.
Header information including information indicating the number of divided images and the sizes of the first and second block portions is generated. Then, in step S31, each block number or address obtained by conversion / generation of the converted second block portion corresponding to the first block portion, the conversion parameter, and the generated header information are decoded. / Multiplexed and transmitted. When step S31 is completed, the process ends.

In this example, as in the preceding example of the present invention, a header having the format shown in FIG. 14 is used. Multiple areas where the screen is divided (as shown in FIG. 2)
Are all equal in size. However, according to the present invention, the area into which the screen is divided or the area defined is not required to be the same size. As an example, the design of the region as shown in FIG. 16 is such that four regions (1-4) are small in size, three additional regions (5-7) are medium in size, and three The last area (8-10) is used as a hierarchical structure having a large size. In order to use such an area definition, an advanced header as shown in FIG. 15 is used. As described above, the header includes information on the size of the screen (horizontal and vertical) and the number of divided screens.
Since the size of the area can be different, so is the hierarchical structure of the blocks indicating the number of layers, the image format indicating the entire format of the image, and information on the size of each block. Therefore, this hierarchical structure shows the definition of a small block and its size, the definition of a middle block and its size, and the definition of a large block and its size. As described above, since the information is encoded in the fixed-length format, when a specific area of the image is accessed, even if the image area has a different size, a part of the bit stream including the information of this area is used. It is possible to jump directly to, thus containing different amounts of information. It is not necessary to read through all data streams as in the prior art. The enhanced header allows more specific image coding schemes for specific images, but still allows direct non-continuous access to portions of the data stream.

Referring to FIG. 11, a decoding device constructed according to the present invention will be described. This encoder can be used to decode information encoded by any of the iterative image transform encoders described above. Storage medium 400, communication network 50
0 or another data input device provides the coded bit stream to the decoding area image control unit 17. The decoding area image control unit 17 receives the encoded bit stream 120 and decoding target area information 128 from an external input. The decoding target area information 128 includes a portion of the image to be decoded,
That is, it is used for controlling the bit stream 120. The demultiplexing / decoding unit 9 that demultiplexes and performs decoding receives the selected part 129 of the coded bit stream 120 and stores the number or address 111 of the part 104 of the first block to be decoded in the image memory unit. 11 and the second block number or address 108 and the screen division information 112 corresponding to the first block portion 104 are provided to the conversion source block reproduction unit 10. The decomposition / decoding unit also provides the conversion parameter 110 to the image conversion / generation unit 6. The source block reproduction unit 10 sends the source unit block 106 to the image conversion / generation unit 6 according to the given number or address of the second block part, and the image conversion / generation unit sends the given conversion parameter. The transformation of the transformation source block 106 is performed using 110. The image memory unit 11 stores a plurality of conversion block parts 1 output from the image conversion / generation unit 6.
07 and the output from the image memory is coupled to the control unit 12 for controlling the continuation of the iterative loop of the decoding device. The region extracting unit 28 extracts a predetermined region from an image to be output from the decoding device.

The processing of the decoding device shown in FIG. 11 will be described.
The storage medium 400, the communication network 500, or another data input device supplies the coded bit stream 120 to the decoding area image control unit 17, which is also provided with decoding target area information 128. The decoding area image controller controls the encoded bit stream 120 according to the decoding target.
And outputs the decoded coded bit stream 129 to the decomposition / decoding unit 9. Therefore, when the determined area of the image to be decoded is input, the decoding area image control unit determines the number of predetermined areas including the part of the image to be decoded. Even if the entire region is not displayed, as described above with respect to the encoding device, the range block of a particular predetermined region can be encoded by a domain block at any position in the same region, so that each displayed whole including even one pixel The region is decoded. Therefore, in order for any part of a particular region to be displayed, the entire region must be decoded. However, regions that do not include the displayed range block need not be decoded by the device. Thus, the target coded bitstream 129 includes data corresponding only to a predetermined region of the coded bitstream to be decoded.

The target coded bit stream 129 is decomposed and fixed-length decoded by the decomposing / decoding unit 9, and the target coded bit stream, that is, the first coded bit stream 129 located in the region indicated by the bit stream containing the information to be decoded and displayed. Number or address 104 of part 104 of one block
Is given to the image memory unit 11. Image division information 112
And the number or address 1 of the portion 106 of the second block
08 is given to the conversion source block reproduction unit 10. The conversion parameters 110 obtained by demultiplexing are supplied to the image conversion / generation unit 6. The conversion source block reproducing unit 10 determines the position of the second block on the screen (based on the number or address 108 of the second block part and the position of the memory of the first block part), The source block 106 is then given to 6 and converted according to the conversion parameters 110.

Thus, the conversion block portion 107 obtained from the image conversion / generation section 6 is supplied to the image memory section 11 and stored in a position according to the setting of the number or address 11 of the first block portion. This procedure is repeated until all decoded target encoded bitstreams have been read and all second transformed block positions 107 have been written to the image memory. Then, the image information is transmitted to the control unit 1 that couples the control signal 116 to the source block reproducing unit.
This is fed back to the source block generator 10 under the control of 2. Therefore, the procedure of the IFS decoding is executed only by the repetition and recursive procedures of the predetermined number of repetitions for the converted second block portion stored in the image memory unit 11. When a predetermined number of repetitions are performed, the decoded image 11 is displayed on the target area extracting unit 28 that selects and displays only the image portion set by the decoded target area information 128.
7 is finally given, and the selected information is output as an output image 150. It should be appreciated that less than the full image needs to be decoded, but more will be decoded than the final display. For example, if the user wants to display the part of the image of the area 1 and the part of the image of the area 2, the entire area of the area 1 and the area 2 is decoded, but the areas 3, 4 and the following areas are decoded. No need to be decrypted.

The decoding area image control unit 1 of the decoding apparatus of the present invention
7 will be described with reference to FIG. As shown in FIG. 12, the decoding area image control unit 17 includes a header reading unit 18 that reads a header from the decoding bit stream 120,
An encoded code reading unit 19 for reading an encoded code including address read information from an encoded bit stream 130 supplied from a header reading unit, and a decoding area of a screen image is determined and decoded target block image information 131 is read. Decoding target area information 1 to be given to address calculating section 21
And a decoding target area determining unit 20 for receiving the decoding target area 28. The read address calculating unit 21 determines a read address of information from a part of the coded bit stream 120 corresponding to a part of the image to be decoded, and supplies read address information 132 to the decoded code reading unit 19.

The operation of the decoding area image control unit 17 will be further described with reference to the flowchart of FIG. In the first step S32, the header reading unit 18 receives the encoded bit stream 120, and extracts header information therefrom. As shown in FIG. 14, and as noted above, this header includes information on screen size, information on the number of screen divisions, information on the size of the first and second blocks. Contains. Otherwise, if the image has a hierarchical structure, the header includes the information shown in FIG. Then, the operation proceeds to step S33, but the decryption target area information 128
Is input to the decoding target area determination unit 20, and determines the area of the image to be decoded, that is, the area of the image including at least one pixel that is finally displayed as an output image. As an example, as shown in FIG. 17, three range blocks Rk, Rk + 1, and Rk + 2 indicated by oblique lines are located in the area 1 and decoded. As described above, these three range blocks (first blocks) are generated by mapping the transformed domain blocks (second blocks) located in the same area 1. Therefore, in order to decode the image data described by the three range blocks Rk, Rk + 1, and Rk + 2, it is necessary to decode all the blocks in only the region 1, but decode the domain blocks in other image regions. It is not required to be done.
Similarly, decoding the four indicated range blocks in region 4 requires that all domain blocks in region 4 only be decoded.

The operation proceeds to step S34, and the decoding target area determined by the decoding target area determining unit 20 including the addresses of the image information of all the range blocks in the area 1 is the decoding of the target area of each range block to be decoded. Are calculated and given to the read address calculating unit 21 which is output as read address information 132. The operation proceeds to step S35,
The decoding / reading unit 19 uses the address information to read out the codes of all the range blocks in the area 1 that is the decoding target area. The operation then proceeds to step S36 (this operation is performed by the decomposer / decoder), where the target encoded bitstream 129 corresponding to the region of the particular image to be decoded
Are decoded, and the decoded data is output from the decomposition / decoding unit 9. As shown in FIG. 14, the encoded code according to the header includes only the number or address 108 of the second block portion, and only the conversion parameter associated with each first block portion. Therefore, by performing fixed-length coding on the number or address of each second block and the conversion parameter on the coding side, the position of the coded bit stream of the coded code of the specific first block is immediately calculated. Is accessed. An example of decryption on a general purpose computer is shown in FIG.
Is described in the flowchart of FIG.

In the first step S41, in FIG. 18, decoding target area information is input, and an encoded bit stream including only information corresponding to an image area to be decoded is extracted based on the decoding target area information. In a succeeding step S42, the coded bit stream extracted in the step S41 and its associated header are demultiplexed and subjected to fixed-length coding, and the number of addresses of the first block portion, The number of addresses of the part, the conversion parameter of each second block part, and the screen division information are decomposed. The operation proceeds to step S43, where a source block portion (second block portion) for positioning the final converted second block is obtained by the conversion parameter obtained in step S42 and the conversion parameter obtained in step S43. Is converted / generated based on the converted source block portion.

The operation then performs a query S to determine whether all first block parts of the decoding target area have been decoded.
Proceed to 46. If all first block parts have not been decoded, operation returns to step S43, where the next original transformed block part for the next first block part is reproduced. On the other hand, if all first block portions have been decoded, the operation proceeds to inquiry S47, which determines whether the iterative transformation operation has been performed a predetermined number of times. Otherwise,
The operation returns to step S43 again, and the above operation is executed again. On the other hand, when the operation proceeds to step S48, the image of the target area that is required to be displayed by the user is extracted from the decoded image, and the extracted image is output as the final decoded image.

The local memory section 116 of the example of the encoding apparatus shown in FIG. 7 and the local memory section of the area repetition function encoding section of the example shown in FIG. 9 will be further described with reference to FIG.

Since these local memory units are frequently accessed, for example, SDRAM (synchronous dynami
c random access memory) is used. Alternatively, a cache memory may be used.

The recording capacity of the local memory is determined by, for example, def_x_range (pixel) in the horizontal direction and def_y_range (row) in the longitudinal direction shown in FIG. If the first block image Rx is the current coded target block, the second block image Dk is the reference image to be searched. In this case, since the number or address of the second block image Dk is searched only for the area determined by the broken line, the number of target blocks is smaller than when the entire screen is used as the search range. As a result, the bit length required to encode the number of addresses of the second block image is reduced. Therefore, by dividing the image into regions to be searched, the search range is narrowed in the encoding device, and the processing time is significantly reduced.

A recording medium for storing a computer program for encoding and decoding according to the present invention includes instructions for setting the encoding technique shown in FIG. 10 and the decoding technique shown in the flowchart in FIG. Specific examples of such media include digital video discs. Examples of applications of the present invention are image transmission / reception devices, image databases,
Image compression / decoding device for downloading image information from the Internet, electronic still camera, game device, encoding / decoding device having display device capable of displaying various sizes, document compression in image editing device Storage unit,
There is a writing device or a fast restoration chip. These are also realized by software.

The transformed second most similar to the first block
In order to select a block, not only threshold processing but also minimum value detection is executed. Thus, the procedure of the second block generates an error that is not only smaller than the error of the predetermined value but also smaller than all other second blocks. However, as another example, only threshold processing needs to be performed. In this case, if an error smaller than the threshold is found, a second block of that error is populated and no further search is performed. The accuracy of the match will decrease, but the processing time will decrease.

In general, according to the present invention, an improved IF
An S encoding and decoding technique is provided. The encoding technique involves dividing an input image into portions of a first block and a second block. Each second block portion is subjected to a conversion process, and which of the converted block portions in the same predetermined area as the input image as the preselected first block portion is selected as the first selected first block portion. Is determined to be closest to the part of the block. Selected second
The block position information indicating the position of the block and the conversion parameter indicating the conversion process of the selected second block are output as codes.

The decoding technique according to the present invention allows decoding from a code representing a block from a predetermined area of an image.
The received code includes block position information indicating the position of a specific area of the image of the block that has been converted in advance, and conversion parameters that indicate the conversion processing of the block that has been converted in advance. In the conversion process of the block identified based on the block position information and based on the conversion parameter of the predetermined region, the predetermined region of the decoded image from the conversion processing block recursively executes the reconstruction.

For setting the encoding and decoding method according to the invention, in the case of encoding, an encoded bit stream,
In the case of decoding, a recording medium for storing a computer program for operating a general-purpose computer for outputting a decoded predetermined area of an encoded image is provided.

[0080]

As described above, the coding technique according to the present invention maps the transformation between range blocks and domain blocks of a region generated by dividing the screen into a plurality of equal or unequal. By dividing the screen into a large number of regions, the size of each image is reduced. As a result, when the area of the image to which the target image to be decoded belongs is small, the decoding time and the storage capacity of the decoding device are reduced.

The decoding technique according to the invention requires that only the coded data corresponding to the block part of the target area to be decoded be read from the bit stream. Therefore,
Unlike the prior art, there is no need to read the preceding entire encoded bit stream, and the decoding processing time is reduced.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating an example of a configuration of an encoding device.

FIG. 2 is a diagram illustrating a mapping process between image blocks in a plurality of predetermined image regions.

FIG. 3 is a flowchart showing a series of steps of an encoding method for encoding a target area of a predetermined image.

FIG. 4 is a diagram illustrating a mapping process between image blocks.

FIG. 5 is a block diagram illustrating an example of a configuration of an encoding device.

FIG. 6 is a flowchart showing a series of steps of an encoding method for encoding a target area of a predetermined image.

FIG. 7 is a block diagram illustrating an example of a configuration of an encoding device.

FIG. 8 is a block diagram illustrating an example of a configuration of an encoding device.

FIG. 9 is a block diagram showing an internal structure of each repetition function coding unit.

FIG. 10 is a flowchart showing a series of steps of an encoding method for encoding a target area of a predetermined image.

FIG. 11 is a block diagram illustrating an example of a configuration of a decoding device that decodes a predetermined image region.

FIG. 12 is a block diagram illustrating an internal configuration of a decoded image area control unit illustrated in FIG. 11;

FIG. 13 is a block diagram illustrating an example of a configuration of a decoding device that decodes a predetermined image region.

FIG. 14 is a diagram illustrating a configuration of an encoded bit stream and a header portion.

FIG. 15 is a diagram illustrating another configuration of an encoded bit stream and a header portion.

FIG. 16 is a diagram showing a block selection structure according to another configuration of the encoded bit stream shown in FIG.

FIG. 17 is a diagram showing a mapping between a range block and a domain block of a region of a predetermined coded image.

FIG. 18 is a flowchart showing a series of steps of a decoding method.

FIG. 19 is a block diagram illustrating a configuration of a conventional encoding device of the related art.

FIG. 20 is a block diagram showing a configuration of a conventional prior art decoding device.

FIG. 21 is a diagram showing a mapping between a range block and a domain block of a local memory unit.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 Encoding device, 2 decoding device, 3 iterative function encoding part, 4 iterative function decoding part, 6 image memory part, 9 search area determination part, 15 image memory part, 140 maximum allowable memory determination part

Claims (31)

[Claims] 1. An encoding method for encoding an image, comprising: dividing an input image into a plurality of first block portions; and dividing the input image into a plurality of second block portions. Performing a conversion process on each of the second block portions to generate a plurality of converted block portions; and performing a conversion process on the same predetermined area of the input image as the selected first block. Determining a portion of the transformed block that is most similar to a portion of the selected first block from among the plurality of transformed blocks located; and determining a portion of the transformed block that is determined. Selecting the corresponding second block portion; and block position information indicating the position of the selected second block portion and the conversion process of the selected second block portion. And outputting a conversion parameter representing the following as a code. 2. The encoding method according to claim 1, further comprising a step of dividing the input image into a plurality of regions, wherein the predetermined region corresponds to one of the plurality of regions. . 3. The method according to claim 1, further comprising the step of generating header information including area information for specifying one of the plurality of areas, and wherein the output step outputs the code together with the header information. 3. The encoding method according to item 2. 4. The method according to claim 1, further comprising the step of storing the image of the predetermined area in a local memory.
Coding method as described. 5. An encoding method for encoding an image, comprising: dividing an input image into a plurality of regions; and performing iterative transform encoding on each of the plurality of regions in parallel. Performing the iterative transformation in parallel; dividing the input image into a plurality of first block portions; dividing the input image into a plurality of second block portions; Performing a conversion process on each second block portion to generate a plurality of converted block portions; and a plurality of converted block portions located in the same predetermined area as the selected first block of the input image. Determining, from the converted blocks, a portion of the converted block that is most similar to the selected portion of the first block; and determining a portion of the converted block corresponding to the determined portion of the converted block. 2 blocks Selecting a portion; and outputting, as codes, block position information indicating the position of the selected second block portion and a conversion parameter representing the conversion process of the selected second block portion. Multiplexing the code generated by each iterative transform coding. 6. A decoding method for decoding encoded data composed of a plurality of codes, comprising: obtaining block position information indicating a position of a source block and iterative transform encoding of a block using only a predetermined area of an image. Receiving the converted parameters, and performing a conversion process based on the converted parameters, in the predetermined area of the image, until a predetermined condition is achieved for each block to be decoded, to a block indicated by the block position information. Performing the conversion process recursively. 7. The decoding method according to claim 6, wherein the predetermined area corresponds to an area obtained by dividing the image into a plurality of areas. 8. The method according to claim 6, further comprising: receiving information indicating an area of the image to be decoded; and extracting a code from an encoded data stream corresponding to the indicated area. The decoding method described. 9. The coded data includes header information including information specifying a plurality of regions obtained by dividing the image, and the conversion process is performed according to the header information. 7. The decoding method according to item 6. 10. In a recording medium on which a plurality of codes are recorded, a step of dividing an input image into a plurality of first block portions, and a step of dividing the input image into a plurality of second block portions. Dividing, a step of performing a conversion process on each second block portion to generate a plurality of converted block portions, and a same predetermined area as the selected first block of the input image. Determining the portion of the transformed block that is most similar to the portion of the selected first block from among the plurality of transformed blocks located at Selecting the portion of the second block corresponding to the following; and block position information indicating the position of the portion of the selected second block; And a step of outputting a conversion parameter representing a conversion process as a code, and a code generated by iterative conversion coding in each step. 11. An encoding apparatus for encoding an image, comprising: a first division unit that divides an input image into a plurality of first block portions; and a plurality of second blocks that divide the input image into a plurality of second blocks. A second dividing unit that divides each of the second block portions into a plurality of converted block portions by performing a conversion process on each second block portion; A determining unit that determines, from among the plurality of converted blocks located in the same predetermined area as the first block, the converted block portion most similar to the selected first block portion; A selection unit that selects the second block part corresponding to the determined converted block part; block position information indicating the position of the selected second block part; and the selected second block part Bro An output unit for outputting, as a code, a conversion parameter representing the conversion process of the block portion. 12. The code according to claim 11, further comprising an area dividing unit that divides the input image into a plurality of areas, wherein the predetermined area corresponds to one of the plurality of areas. Device. 13. The information processing apparatus according to claim 1, further comprising a header information generating unit configured to generate header information including area information specifying one of the plurality of areas, wherein the output unit outputs the code together with the header information. The encoding device according to claim 12, wherein 14. The apparatus according to claim 11, further comprising a local memory unit for storing the image of the predetermined area.
An encoding device according to claim 1. 15. An encoding method for encoding an image, comprising: an area determining unit that divides an input image into a plurality of areas; and a plurality of encodings that perform iterative transform encoding on each of the plurality of areas in parallel. A plurality of encoding units for performing the iterative transform encoding in parallel, a first division unit for dividing an input image into a plurality of first block portions, A second dividing unit that divides the image into a plurality of second block parts, a converting unit that performs a conversion process on each of the second block parts to generate a plurality of converted block parts, A portion of the converted block most similar to the portion of the selected first block from among the plurality of converted blocks located in the same predetermined area as the selected first block of the selected image And a determining unit for determining A selecting unit that selects a part of the second block corresponding to the part of the converted block; block position information indicating a position of the part of the selected second block; An encoding apparatus, comprising: an output unit that outputs a conversion parameter representing the conversion processing of a part as a code; and a multiplexing unit that multiplexes a code generated by each repetitive transform coding. 16. A decoding device for decoding encoded data composed of a plurality of codes, comprising: a block position information indicating a position of a transformation source block; and a block obtained by iterative transform encoding of a block using only a predetermined area of an image. A receiving unit that receives the converted parameters, and a block indicated by the block position information until a predetermined condition is achieved for each block to be decoded in the predetermined area of the image by a conversion process based on the conversion parameters. And a conversion unit for performing a conversion process recursively. 17. The decoding device according to claim 16, wherein the predetermined area corresponds to an area obtained by dividing the image into a plurality of areas. 18. The image processing apparatus further comprising: an area determining unit that receives information indicating an area of the image to be decoded; and an extracting unit that extracts a code from an encoded data stream corresponding to the indicated area. The decoding device according to claim 16. 19. The coded data includes header information including information specifying a plurality of regions obtained by dividing the image, and the conversion process is performed according to the header information. 17. The decoding device according to item 16. 20. An encoding apparatus for encoding an image, comprising: a first division unit configured to divide an input image into a plurality of first block portions; and a plurality of second blocks configured to divide the input image into a plurality of second blocks. A second dividing unit that divides each of the second block portions into a plurality of converted block portions by performing a conversion process on each of the second block portions; Determining means for determining, from among the plurality of converted blocks located in the same predetermined area as the first block, the converted block portion most similar to the selected first block portion; Selecting means for selecting the second block portion corresponding to the determined converted block portion; block position information indicating the position of the selected second block portion; Output means for outputting, as a code, a conversion parameter representing the conversion processing of the second block portion. 21. The code according to claim 20, further comprising an area dividing unit for dividing the input image into a plurality of areas, wherein the predetermined area corresponds to one of the plurality of areas. Device. 22. The image processing apparatus further comprising: header information generating means for generating header information including area information specifying one of the plurality of areas, wherein the output means outputs the code together with the header information. 22. The encoding device according to claim 21, wherein 23. The apparatus according to claim 2, further comprising a local memory means for storing the image of the predetermined area.
0. The encoding apparatus according to claim 0. 24. An encoding apparatus for encoding an image, comprising: area determining means for dividing an input image into a plurality of areas; and a plurality of encoding apparatuses for performing iterative transform encoding on each of the plurality of areas in parallel. A plurality of encoding means for performing the iterative transform encoding in parallel, a first dividing means for dividing an input image into a plurality of first block portions, A second dividing unit that divides the image into a plurality of second block portions, a converting unit that performs a conversion process on each of the second block portions to generate a plurality of converted block portions, A portion of the converted block most similar to the portion of the selected first block from among the plurality of converted blocks located in the same predetermined area as the selected first block of the selected image Means for determining Selecting means for selecting the second block portion corresponding to the converted block portion, block position information indicating the position of the selected second block portion, and the selected second block portion. An encoding apparatus, comprising: output means for outputting, as a code, a conversion parameter representing the conversion processing of a block portion; and multiplexing means for multiplexing a code generated by each repetitive transform coding. 25. A decoding device for decoding encoded data composed of a plurality of codes, comprising: a block position information indicating a position of a source block and repetitive transform encoding of a block using only a predetermined area of an image. Receiving means for receiving the converted parameter, and a block indicated by the block position information until a predetermined condition is achieved for each block to be decoded in the predetermined area of the image by a conversion process based on the conversion parameter. And a conversion means for performing a conversion process recursively. 26. The decoding apparatus according to claim 25, wherein the predetermined area corresponds to an area obtained by dividing the image into a plurality of areas. 27. An image processing apparatus, further comprising: area determining means for receiving information indicating an area of the image to be decoded; and extracting means for extracting a code from an encoded data stream corresponding to the indicated area. The decoding device according to claim 25. 28. The coded data includes header information including information specifying a plurality of regions obtained by dividing the image, and the conversion process is performed according to the header information. 26. The decoding device according to item 25. 29. Each code is composed of a plurality of codes including block position information indicating the position of each source block and conversion parameters obtained by iterative transform coding of a block using only a predetermined area of an image. A recording signal on which a recording signal containing encoded data is recorded, the recording medium being decodable by a decoding device, wherein the block position information contained in the encoded data and the conversion parameter also contained in the encoded data are Receiving, by the conversion process based on the conversion parameter, recursively perform the conversion process on the block indicated by the block position information until a predetermined condition is satisfied for each block to be decoded in the predetermined region of the image. And a recording signal decoded by each step of performing the step of performing That recording medium. 30. A recording medium on which an encoding program is recorded, wherein an input image is divided into a plurality of first block portions, and the input image is divided into a plurality of second block portions. Dividing, a step of performing a conversion process on each second block portion to generate a plurality of converted block portions, and a same predetermined area as the selected first block of the input image. Determining a portion of the transformed block that is most similar to a portion of the selected first block from among the plurality of transformed blocks located at Selecting the portion of the second block corresponding to the following; and block position information indicating the position of the portion of the selected second block and the position of the selected portion of the second block. And a step of outputting a conversion parameter representing the conversion process as a code. 31. Each code is composed of a plurality of codes including block position information indicating the position of each source block and conversion parameters obtained by iterative transform coding of a block using only a predetermined area of an image. Receiving, on a recording medium on which a decoding program for decoding encoded data to be decoded, block position information included in the encoded data and a conversion parameter also included in the encoded data, a conversion based on the conversion parameter Performing a conversion process recursively on a block indicated by the block position information until a predetermined condition is achieved for each block to be decoded in the predetermined area of the image by the processing. A recording medium on which a decryption program is recorded.