JP2796216B2 - Image encoding method and image decoding method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image encoding method in image data processing, and more particularly to an image encoding method capable of reducing the amount of image data while minimizing the deterioration of the image quality of a printed image, and an image decoding method. The present invention relates to a chemical treatment method.

[0002]

2. Description of the Related Art Generally, the amount of image data in the printing field is enormous compared to a still image displayed on a television monitor. For example, the amount of image data in the latter is about 1 MB, whereas the amount of image data in the former is about 1 MB. A4 size manuscript 6
It reaches as much as 0 MB (400 lines / inch).

[0003] In order to store such a large amount of data as a database as it is, a huge memory is required, and the time required for data transmission becomes very long.

In order to cope with this, an encoding technique for reducing the amount of information of an image, that is, an image data compression technique has been studied, and a compression method such as an orthogonal transform encoding method or a vector quantization method has been conventionally developed. Have been. Among them, the orthogonal transform coding method is particularly employed in the international standard compression method for natural still images.

[0005] These compression methods are so-called irreversible encoding methods, and although a high compression ratio can be expected, they have a drawback that they do not completely return to the original image data even if they are expanded.

In particular, in the case of commercial print images, image quality (for example, smoothness of women's skin and sharpness of contours) is often required rather than image content itself. If an irreversible encoding method such as orthogonal transformation is applied to the image data and the image data is compressed at a high compression rate, the image quality deteriorates and the commercial value of the image is impaired.

[0007]

For this reason, there is a need for a study of an encoding method which saves visually necessary information as much as possible while reducing the amount of image data, and minimizes image quality deterioration due to compression of the relevant portion. Despite this, there was no effective image processing method.

An object of the present invention is to provide an image encoding method and an image decoding method for maintaining a clear image of a given object while keeping the amount of image data low irrespective of the type of image. The purpose is to provide.

[0009]

[0010]

Means for Solving the Problems To achieve the above object,
Therefore, in the image encoding processing method according to the first aspect of the present invention,
A method of encoding the pixel data to compress image data, comprising: (a) setting a contour of a predetermined target object on an image to form contour data; and (b) forming a plurality of the image data. And (c) dividing the pixel block based on the contour data,
An outline block group including the outline data, an object block group not including the outline data and including the object image, and a non-object that is obtained by removing the outline block group and the object block group from all pixel blocks. (D) image data in a pixel block belonging to the outline block group based on the outline data, and excludes the object image data including the object image and the object image. Classifying into non-target image data, and replacing the target image data with non-target image data to obtain image data of a pseudo non-target block group; (e) belonging to the target block group The first encoding process with a relatively low compression ratio is performed on the pixel blocks, and the second encoding process with a relatively high compression ratio is performed on the pixel blocks belonging to the non-target block group. That is, the pixel blocks belonging to the contour block group have a relatively low compression ratio of the third block.
A step of obtaining a coded pixel data for the plurality of pixel blocks by performing a coding process on the pixel blocks belonging to the pseudo non-object block group by performing a fourth coding process with a relatively high compression rate. And the following.

An image decoding method according to a second aspect of the present invention is a method for decoding image data encoded by the method of the first aspect, wherein: The first decoded data is obtained by performing a first decoding process corresponding to the first encoding process on the encoded image data, and the second decoding process is performed on the image data encoded by the second encoding process. By performing a second decoding process corresponding to the process, a second decoded data is performed, and a third decoding process corresponding to the third coding process is performed on the image data coded by the third coding process. Thus, the third decoded data is converted to the fourth data with respect to the image data encoded in the fourth encoding process.
A step of obtaining fourth decoded data by performing a fourth decoding process corresponding to the encoding process; and (b) converting the third decoded data and the fourth decoded data based on the contour data into object image regions, respectively. Generating modified decoded data by combining the third decoded data of the target image region with the fourth decoded data of the non-target image region; and (c) the modified decoding. Combining the data, the first decoded data, and the second decoded data to obtain decompressed image data of the image data.

[0012]

[0013]

[0014]

[0015]

[Action] image encoding method according to claim 1, first, the image to set the outline of the predetermined object, thereby forming the image data including the contour as the contour data, the image data into a plurality of pixel blocks To divide.

Based on the contour data, the plurality of pixel blocks are divided into a contour block group including the contour data,
The object blocks are grouped into an object block group that does not include the contour data but includes the object image and a non-object block group that excludes these from all pixel blocks.

Next, the image data of the pixel blocks belonging to the outline block group is classified into object image data and non-object image data based on the outline data, and the object image data is replaced with the non-object image data to generate a pseudo image. Obtain image data of the non-object block group.

Of these pixel blocks, the first encoding process with a relatively low compression ratio is performed on the pixel blocks belonging to the object block group, and the first encoding process with a relatively high compression ratio is performed on the pixel blocks belonging to the non-object block group. The second encoding process is performed by a third encoding process with a relatively low compression ratio for the pixel blocks belonging to the contour block group, and the fourth encoding process is performed with a relatively high compression ratio on the pixel blocks belonging to the pseudo non-object block group. Encoding processing is performed respectively. As a result, it is possible to minimize the deterioration of the image quality of the target object image and its outline while keeping the data amount of the entire image small . Further, in the image decoding method according to claim 2, it has a method suitable for decoding the image data encoded in claim 1 of the image encoding processing method, first of all, the second , The image data encoded in the third and fourth encoding processes are respectively converted into first,
By performing the second, third, and fourth decoding processes, first, second, third, and fourth decoded data are obtained.

Both the third decoded data and the fourth decoded data are decoded data of the pixel blocks belonging to the outline block group, and these decoded data are converted into the image data of the object image area and the non-object image area based on the outline data. By dividing the third decoded data of the target image region and the fourth decoded data of the non-target image region,
Obtain modified decoded data. Since the data of the target image area in the modified decoded data is obtained by decoding the image data encoded at a relatively low compression rate, the deterioration of the image quality can be minimized.

Decompressed image data is obtained by combining the modified decoded data, the first decoded data, and the second decoded data.

[0021]

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a block diagram showing an embodiment of an image coding method and an image decoding method according to the present invention; Of course not.

This embodiment will be described in the following order.

(A) Creation of contour data (B) Blocking of image data (C) Classification of pixel blocks (D) Creation of pseudo non-object blocks (E) Image compression method (F) Image restoration method (G) Example of Apparatus for Implementing Method According to the Present Example (H) Modification Example

(A) Creation of Contour Data In this embodiment, for the sake of explanation, image processing on a simple clock face image as shown in FIG. 7A will be considered.

When the operator wants to set the clock face 41 of the original image 40 as an object, the image data corresponding to the outline 41a displayed on the television monitor is all set to "1", and other data are set. The image data is set to “0”, whereby the contour data R shown in FIG. 7B is obtained.

Such contour data R can be easily obtained by, for example, a known contour cutting device. In order to create the contour data R by the contour cutting device, the original image 40 is displayed on a television monitor, and the operator moves the cursor displayed on the screen along the contour line 41a with the mouse at hand. Let it.

The trajectory of the cursor movement is converted into an xy coordinate plane in the pixel matrix, and the contour data value corresponding to the contour is set to "1" and the other contour data values are set to "0". By doing so, contour data R can be obtained.

The contour data R created by assigning “0” and “1” in this way is compressed by vectorization and stored in the memory unit.

(B) Blocking of Image Data In the present invention, image data obtained from the original image 40 is divided into pixel blocks in order to classify image data into blocks. FIG. 8 schematically shows the state of the pixel block division.

As shown in FIG. 8A, the image data is (H
The pixel is divided into (M × N) pixel blocks Bxy each including a plurality of (V) pixels P and each having a plurality of pixels.

FIG. 8B shows one pixel block B.
FIG. 3 is a conceptual diagram showing a configuration of xy, where a pixel block Bxy is
It is a pixel matrix composed of (I × J) pixels, and image data fij is obtained for each pixel.

In the example of FIG. 8B, one pixel block Bxy is composed of (8 × 8) pixels.
× 4) pixels or (16 × 16) pixels.

(C) Classification of Pixel Block Next, the plurality of pixel blocks are divided into an outline block group including outline data and an object including the object image in a closed area formed by the outline block group. The blocks are classified into a block group and a non-object block group obtained by removing the contour block group and the object block group from all the pixel blocks, and the following code table is prepared as preparation for the classification.

(C-1) Creation of Object Block Table First, the contour data R which has been vectorized and compressed and stored in the above (A) is restored by performing a decoding process corresponding to the vectorization. Contour data R obtained by restoration
Corresponds to a large number of (H × V) pixels P as in the image data of the original 40, and as described in FIG.
Or it takes the value of 0. Therefore, the contour data R is divided into (M.times.N) blocks in the same manner as in the block formation of the image data in (B) to form a contour data block Dxy corresponding to the pixel block Bxy.

Further, a code table CT (x, y) of contour data having (M × N) areas corresponding to the (M × N) divided contour data blocks Dxy.
Prepare

The procedure for creating the code table CT (x, y) will be described with reference to the flowchart of FIG.

First, the processing is started from the outline data block D00 with y = 0 and x = 0 (steps S31 and S3).
2).

If at least one of the plurality of outline data included in the outline data block Dxy is "1", it can be determined that the outline is included in the outline data block Dxy. The value of the code table CT (x, y) corresponding to the contour data block Dxy is set to "1" (steps S33, S34).

If all of the plurality of outline data included in the outline data block Dxy are "0", no outline is included in the outline data block Dxy. The value of the code table CT (x, y) corresponding to the block Dxy is set to "0" (step S35).

Next, the variable x is incremented by "1". If the value is less than M, which is the number of pixel blocks in the x coordinate direction, the processing on the y row has not been completed.
Returning to step S33, processing is performed on the next contour data block Dxy (steps S36 and S37).

If the variable x is M in step S37
If so, the operation on the y-th row has been completed. Therefore, “1” is incremented to the variable y, and if the value of the variable y at this time is smaller than N, which is the number of pixel blocks in the y-coordinate direction. For example, the process returns to step S32 to perform an operation on the contour data block Dxy of the next row (step S3).
8, S39). If the value of the variable y is equal to or greater than N, the process ends.

In this way, the process of setting “0” or “1” to CT (x, y) of the code table by the contour data is performed from the upper leftmost contour data block D00 to the lower rightmost one. By performing the processing up to the outline data block D (M-1) (N-1), the code table 42 corresponding to the outline data R as shown in FIG. 9A is completed.

Next, based on the code table 42 corresponding to the contour data, the pixel blocks Bxy of the divided image data are divided into a pixel block group including a contour (hereinafter, referred to as a "contour block group") and a contour block. A pixel block group (hereinafter, referred to as an “object block group”) in a closed region formed by the group and including an object image, and a pixel block group excluding the outline block group and the object block group (hereinafter, referred to as “object block group”). , “Non-target block group”). A code table CT (x, y) is created.

The creation of the code table is basically achieved by performing a “painting” operation on the area of the object image in the closed area formed by the outline block group.

Various methods have been devised for this "painting" process in the field of image processing. For example, "Electronic Information and Communication Handbook Second Volume, edited by the Institute of Electronics, Information and Communication Engineers" (Ohmsha, March 1988) Issued on March 30)
On page 1868.

A code table newly created by replacing the object area in the outline block group with “2” by the “painting” process is the code table 43 shown in FIG. Is data accompanying image data, and this code table is hereinafter referred to as a "target block table" in the sense of a code table for determining a target block or the like.

(C-2) Classification of Pixel Blocks The object block table 4 formed as described above
3, it is determined whether the corresponding pixel block Bxy is a contour block, a target block, or a non-target block based on the value of each data CT (x, y).

That is, the pixel blocks are sequentially read from the memory unit, and if the data CT (x, y) of the object block table 43 corresponding to the pixel block Bxy is "1", the pixel block Bxy is a contour block. Yes,
Similarly, if CT (x, y) is “2”, the corresponding pixel block Bxy is an object block, and
If (x, y) is “0”, the corresponding pixel block B
xy can be determined to be a non-object block.

By performing such a determination for all the pixel blocks to form the outline block group, the object block group, and the non-object block group, it is possible to perform different encoding processing for each pixel block group. become.

(D) Creation of Pseudo Object Block Next, the image data of the pixel block classified as the outline block group in (C) is corrected to create a pseudo non-object block group.

Such a pseudo non-object block is formed by replacing the image data of the object area in the contour block with the appropriate image data of the non-object area. Depends on the procedure.

(D-1) Creation of Object Pixel Table FIG. 11A shows contour data including a contour portion in a contour data block Dxy obtained by dividing the contour data R into (M × N) pieces. FIG. 9 schematically shows a state of a data array of a block Dmn.

As described in (A) above, the portion having the contour data value "1" indicates the contour portion, and the portion having the contour data value "0" indicates the other portion.

The "0" on the object area side from the contour "1" is all replaced with "1" by the "painting" method described above.

On the other hand, a code table DTmn of contour data having (I × J) areas corresponding to the contour data block Dmn having (I × J) pixels is prepared, and the “painting” process is performed. By setting “0” or “1” in the data DTmn (i, j) at the corresponding position in the code table DTmn based on the data value of the contour data block Dmn obtained by the processing shown in FIG. The table DTmn is formed.

The newly created code table DT
By determining whether the value of each data DTmn (i, j) of "mn" is "1" or "0", each image data in the pixel block Bmn corresponding to the position of the data is converted into an object image including an outline. It can be easily determined whether the region belongs to the region of the non-object image or the region of the non-object image.

This code table DTmn is hereinafter referred to as an "object pixel table" in the sense of determining whether the image data of each pixel in the pixel block Bmn belongs to the object area or the non-object area.

(D-2) Creation of Pseudo Non-Object Block The object pixel table DTm created as described above.
The state of the arrangement of each image data in the contour block Bmn corresponding to n is, for example, as shown in FIG. In the figure, the object pixel table D
A boundary line 33 is drawn at the boundary between the pixel representing the contour and the pixel in the non-target area so that the correspondence with Tmn can be easily understood.

The upper part of the boundary line 33 is the image data of the non-target area (hereinafter, simply referred to as “non-target image data”), and the lower part is the image data of the target area including the outline (hereinafter, referred to as “non-target image data”). Simply referred to as “object image data”).

The value of the object image data is replaced in the j-coordinate direction by the value of the non-object image data adjacent to the boundary 33, thereby modifying the contour block MBmn as shown in FIG. Get.

The outline block obtained by replacing the object image data with predetermined non-object image data is hereinafter referred to as a "pseudo non-object block" in the sense that it is similar to the non-object block.

Such a pseudo non-object block MBmn
Is created specifically by the object pixel table DTmn.
Is scanned from the j-coordinate direction for each column, and when the data DTmn (i, j) changes from “0” to “1”, the data “0” of the immediately preceding “0” is read. This can be achieved by storing the image data fij of the pixel block Bmn corresponding to the position (i, j) and replacing the image data below it with this value. However, in the example shown in FIG. 12A, the rightmost column is the object image data, and there is no non-object image data adjacent in the j-coordinate direction. (I-
2) The replacement is performed using the image data of (1).

The creation of these object pixel tables DTmn and the corresponding pseudo non-object blocks MBmn are as follows.
This is performed for all pixel blocks belonging to the outline block group.

In this embodiment, the object image data is replaced by the non-object image data adjacent to the boundary 33 of each column of the contour block in the j coordinate direction. The object image data may be used, and in some cases, all of the object image data may be replaced with an average image data value of the non-object region.

However, when the replacement is performed using the non-object image data as close as possible to the boundary 33, the error from the actual non-object image data when the image is restored can be reduced, and the natural contour A restored image having a portion is obtained.

(E) Image Compression Method The outline block group, the object block group, the non-object block group classified based on the object block table 43 as described above, and the outline block group based on the object pixel table DTmn. The image data in each of the pseudo non-object block groups formed by correcting is subjected to an encoding process as follows.

First, the pixel blocks Bxy are sequentially called from the image data. If the value of CT (x, y) in the corresponding object block table 43 is "1", it is determined that the pixel block Bxy is a contour block. As shown, orthogonal transform coding, for example, discrete cosine transform, is performed, and the first coding process is performed at a low compression ratio ra in which data of most frequency components is stored (compression A). In this case, low compression ratio means
A compression ratio of 1, that is, a case where no compression is performed is also included.

When the value of CT (x, y) in the object block table 43 corresponding to the pixel block Bxy is "2", it indicates that the corresponding pixel block Bxy is the object block. Using the same orthogonal transform coding method as above, the second coding processing is performed at a low compression ratio rb in which data of most frequency components is stored (compression B).

If the value of CT (x, y) in the object block table 43 corresponding to the pixel block Bxy is "0", it indicates that the corresponding pixel block Bxy is a non-object block. The image data is subjected to a third encoding process at a high compression ratio rc that preserves only low-frequency components even in orthogonal transform encoding (compression C).

Also, the pseudo non-object block group created by modifying the contour block group has a high compression ratio rd.
Performs a fourth encoding process (compression D).

The compression ratio ra in the first encoding process
And the compression ratio rb in the second encoding process may be the same, but if the compression ratio ra is set slightly lower than the compression ratio rb, the outline becomes clearer when the image is restored. Thus, a clearer image can be given to the observer.

For the same reason, the compression ratio rc in the third encoding process and the compression ratio rd in the fourth encoding process are described.
Since the compressed pseudo non-object block group is restored and used to correct the image data of the contour block group (described later (F)), the compression ratio rd is equal to the compression ratio rc.
It is desirable to set it equal to or slightly lower than this.
Therefore, the relationship between these compression ratios is ra ≦ rb <rd ≦ rc
It is said.

Of course, the magnitude relation and the degree thereof may be changed as necessary according to the type of the original image, the presence or absence of the purpose of the special effect in the restored image, and the like.

In the above-described encoding processing of the image data belonging to each block group, the position data (x, y) of each pixel block is also attached as an address.

It is preferable that the object pixel table DTmn is also compressed. In this case, since the object pixel table DTmn is binary data (2 bits) of “0” and “1”, a lossless encoding method, for example, MR (Modified Read)
Although the compression process is performed using a code (compression D), it is needless to say that the data may be compressed by another reversible code, for example, a run-length compression method.

The image data compressed as described above is stored in a different memory unit for each pixel block group, and the object pixel table DTxy is also stored in a different memory unit from the compressed image data. These compressed data are made into a database or transmitted via a communication line.

In this way, by performing different encoding processing for each area based on the target object block table 43, while maintaining the image quality of the important target image, the non-important parts based on the initial contour setting Since a high compression ratio can be applied, image data can be largely reduced as a whole. Further, by creating a pseudo non-object block group, it becomes possible to effectively correct the outline portion when an original image is restored as described later.

(F) Image Restoration Method The following procedure is used to restore image data compressed in this manner (hereinafter, compressed image data is simply referred to as “compressed image data”).

(F-1) Decompression of Compressed Image Data As described above, the compressed image data is stored in different memory units for each pixel block group, so that the compressed image data is stored in each memory unit for each pixel block. If the decoding is performed by reading the data, it is not necessary to perform the decoding while determining to which block group the compressed image data belongs.

That is, the compressed image data is called for each pixel block from the memory unit in which the compressed outline block group is stored, and is subjected to a decoding process corresponding to the inverse conversion of the compression A and expanded (decompression A). ), The compressed image data is called for each block from the memory unit in which the compressed object block group is stored, and the decoding process corresponding to the inverse conversion of the compression B is performed to expand the data (decompression B). In addition, the compressed image data is called for each pixel block from the memory unit in which the compressed non-object block group is stored, and decompression is performed by a decompression process corresponding to the inverse conversion of the compression C (decompression C). The compressed image data is called for each pixel block from the memory unit in which the obtained pseudo non-object block group is stored, and restoration is performed by a decompression process corresponding to the inverse conversion of the compression D (decompression D) (F-2) ) Creation of Modified Contour Block Group The image data of the restored contour block group includes image data of an object region including a contour (object image data) and image data of a non-object region (non-object image). Data). Since the outline block group is compressed at a low compression ratio as described above, the restored image is sharp, but on the other hand, the image data of the non-object block group has been subjected to a high compression ratio compression process. When the density difference is poor and the image data is synthesized and restored in pixel block units, the non-object image data of the contour block group and the non-object image data of the non-object block group adjacent thereto are the former. Is clearly restored and the latter is restored in a deteriorated state, so that a clear density step occurs between both pixel blocks. This gives the observer the impression that the outline has oozed outward. In order to eliminate the "smearing" of the outline, the following correction is made to the decoded image data of the outline block group to form a corrected outline block group.

That is, first, each of the encoded object pixel tables DTmn is restored by a decoding process corresponding to the encoding process. The object pixel table DT of the decoded pseudo non-object block group
The pseudo non-target block MBmn corresponding to mn is called, and the non-target image data is extracted from the image data belonging to the pseudo non-target block MBmn based on the target pixel table DTmn.

In this extraction operation, the data DTmn (i, j) (FIG. 11A) of the object pixel table DTmn is sequentially scanned, for example, for each column from the j-coordinate direction. In this case, this can be easily achieved by sequentially storing the image data of the pixel block Bxy at the position corresponding to the data in the first memory unit.

An object pixel table DT is obtained in a similar manner.
Based on mn, target image data is extracted from the image data of the corresponding pixel block Bmn of the decoded outline block group, and is sequentially stored in the second memory unit.

Finally, the non-object image data of the pseudo non-object block MBxy stored in the first memory unit and the second
Are synthesized based on the respective position data (i, j) with the object image data of the pixel block Bmn of the contour block group stored in the memory unit of (1) to form a corrected contour block.

This correction operation is performed on the pixel blocks Bxy belonging to all the outline block groups to form a corrected outline block group.

Since the non-object image data in this group of modified contour blocks is formed from a pseudo non-object block stored at a high compression ratio rd similar to the compression ratio rc of the non-object block, it is restored. Even when the images are synthesized, the sharpness and the density can be reproduced almost in the same manner as the adjacent non-object block, and the above-described “smearing” of the contour does not occur. Also, the compression ratio rc and the compression ratio rd are set to have substantially the same high compression ratio in principle, or set so that the latter is slightly lower than the former. Is made variable, finer image adjustment at the contour can be performed, and "bleeding" of the contour can be further reduced.

The image data of the corrected contour block group, the object block group, and the non-object block group thus obtained are synthesized based on the position data (x, y) attached to each pixel block. And finally restore the entire image.

At this time, if the image in the background area is subjected to smoothing processing, for example, using a 3 × 3 smoothing matrix, and the data degraded by the block distortion is corrected, the image in the non-object area becomes smooth. .

The original image is usually a color image in many cases, but in the case of such a color image, the color-separated yellow (Y), magenta (M), cyan (C), and black (K) Or the above-described image processing is performed on components such as red (R), green (G), blue (B), or color variables after color coordinate conversion, such as luminance (Ys) and color difference (Is, Qs). Finally, these are combined and restored.

(G) Example of Apparatus for Implementing Method According to the Present Embodiment Next, an example of an image processing apparatus for executing the above-described image compression method and image restoration method will be described.

This image processing device comprises an image data compression device and an image data decompression device, both of which may be configured integrally or may be independently configured and arranged at different locations. There is also.

(G-1) Image Data Compression Apparatus FIG. 1 is a block diagram showing the configuration of the image data compression apparatus 100, and FIG. 3 is a flowchart showing the operation of the image data compression apparatus 100.

First, contour data for the original image is obtained from an external contour creating device, and an object block table 43 is created in the object block table creating section 3 based on the outline data (steps S1, S2, For details of creating a block table, see (C
-1)).

The information of the contour data R is also input to the object pixel table creation unit 4. The object pixel table creating unit 4 classifies the object image data and the non-object image data in each outline block Bmn based on the outline data R and the object block table. A code table (object pixel table DTmn) having a corresponding area is created (Step S3, for details of the method of creating the object pixel table DTmn, refer to the above (D-1)).

On the other hand, the pixel block dividing section 2 divides the image data of the original image 40 stored in the image memory section 1 into a plurality of pixel blocks Bxy (step S4; B)). In the pixel block determining unit 5, the pixel block dividing unit 2
The image data divided into a plurality of pixel blocks Bxy is called for each pixel block to perform a discrimination process.

This pixel block discrimination processing is based on the data CT (x, y) of the object block table 43,
First, block determination processing is started from the upper left pixel block B00 (steps S5 and S6), and it is determined whether the value of CT (x, y) corresponding to the pixel block Bxy is "1" (step S7). ).

The image data compression apparatus 100 includes first to fifth compression units 6, 7, 8, 9, and 10, and the value of CT (x, y) corresponding to the pixel block Bxy is "1". "
Then, since the pixel block Bxy is a contour block,
The compressed image data is sent to the first compression unit 6 and subjected to an encoding process (compression A) with a low compression ratio ra.
(steps S8 and S9). The image data of the contour block is stored in the pseudo non-object block creating unit 1.
1, the pseudo non-object block creating unit 11
Reads out the object pixel table DTmn corresponding to the contour block Bmn from the object pixel table creating unit 4 and, based on the object pixel table DTmn, converts the target object image data of the outline block into predetermined non-object image data. To form a pseudo non-object block MBmn (step S10, see (D-2) for details of creating the pseudo non-object block MBmn).

The pseudo non-object block MBmn is sent to the fourth compression unit 9 and subjected to an encoding process (compression D) at a high compression rate rd, and the memory unit 12 of the compressed image storage unit 12
d (steps S11 and S12).

If it is determined in step S7 that the value of CT (x, y) is not "1", the next step S1
3 to determine whether the value of CT (x, y) is “2”. If the value is “2”, the pixel block Bxy is an object block, and is sent to the second compression unit 7. Low compression ratio r
b, and performs an encoding process (compression B), and stores it in the memory unit 12b of the compressed image storage unit 12 (Steps S14 and S14)
15).

If it is determined in step S13 that the value of CT (x, y) is not "2", the pixel block B
Since xy is a non-target block, it is sent to the third compression unit 8 to perform encoding processing (compression C) with a high compression ratio rc,
The compressed image is stored in the memory unit 12c of the compressed image storage unit 12 (steps S16 and S17).

In this way, one pixel block is determined, and when the processing of compression A, compression B, compression C, or compression D is completed, the variable x is incremented by "1", and the variable x If the value is less than M, the process returns to step S7, and the processes from step S7 to step S17 are repeated for the next pixel block Bxy (steps S18 and S19). Therefore, the variable y is incremented by “1” and the process proceeds to the determination of the next row, and the same processing is repeated until the variable y becomes N. When the variable y becomes N, it means that the encoding process has been completed for all the pixel blocks (steps S20 and S21).

Finally, the data DTmn (x, y) of the object pixel table DTmn is compressed (compressed E) by MR encoding in the fifth compression section 10 and stored in the memory section 12 e of the compressed image storage section 12. The image data is saved (steps S22 and S23), and the image data compression operation ends.

The compressed image data obtained in this manner is stored in a data state in which each color component of the color-separated image data can be distinguished from each other, and is stored for each pixel block Bxy. The position data (x, y) of the pixel block Bxy is stored together with the image data, and an image can be synthesized according to the address at the time of image restoration described later.

The compressed image data and the compression object pixel table data are collectively referred to as “compressed data”. (G-2) Image Data Decompression Device FIG. 2 is a block diagram of the configuration of the image data decompression device 200, and FIG. 4 is a flowchart showing compressed data decompression operation in the image data decompression device 200.

The image data decompression device 200 includes a compressed image storage unit 13 for compressed data.
When the compressed image storage unit 12 is integrated with or close to the image data
00 may be shared with the compressed image storage unit 12.

The compressed image storage unit 13 has a memory unit 13a, 13b, 13c, 1 corresponding to the compressed image storage unit 12.
3d and 13e, which are compressed image data of the contour block group, image data of the object block group, image data of the non-object block group, image data of the pseudo non-object block group, and the object, respectively. Pixel table data is stored.

These memory units 13a, 13b, 13
c, 13d, the compressed data of each block group is read out in units of pixel blocks.
The compressed image data of the outline block group is read out in units of pixel blocks from a, and the first decompression unit 14 performs a decoding process corresponding to the inverse conversion of compression A to decompress (decompression A) (step S24).

Next, the compressed image data of the object block group is read out from the memory unit 13b in units of pixel blocks, and the second decompression unit 15 performs a decoding process corresponding to the inverse conversion of the compression B, and Extend (extend B) (step S25).

The compressed image data of the non-target block group is read out from the memory unit 13c in pixel block units, and the third decompression unit 16 performs a decoding process corresponding to the inverse conversion of the compression C to perform decompression. (Extension C) (step S26).

Further, the compressed image data of the pseudo non-object block group is read out from the memory unit 13d in pixel block units, and the fourth decompression unit 17 performs a decoding process corresponding to the inverse conversion of the compression D. Extend (extend D) (step S27). Then, the compressed target pixel table DTmn is read from the memory unit 13e, and the fourth decompression unit 18 performs decompression processing corresponding to the inverse conversion of the compression E to restore it, and stores it in the memory unit 19 (step S28). ).

The modified contour block creating section 20 stores the image data of the contour block group decoded by the first decompression section 13 and the image of the pseudo non-object block group decoded by the third decompression section 15. Data is given, and the object pixel table DT stored in the memory unit 17
While sequentially reading mn, the data DTmn (x, y)
The target image data of each of the contour blocks and the non-target image data of the corresponding pseudo non-target block are extracted based on the values of the contour blocks, and the two are combined to create a corrected contour block group (step S29: corrected contour block). The detailed description of the group creation is described above (F-2).

When the image data is color image data, the above processing is performed for each color component.

In this way, the second and third extending portions 15,
The image data of the object block group and the non-object block group restored in step 16 and the modified contour block group formed by the modified contour block creation unit 20 are combined with the image data of the next stage synthesis unit 21.
Sent to The synthesizing unit 21 synthesizes the restored image data of each of the pixel blocks of the object, the non-object, and the contour with reference to the position data (x, y) attached to each pixel block, thereby obtaining a final image. First, an image corresponding to the original image is obtained. When the image data is color image data, the above-described synthesis of the image data is performed for each color component.

The synthesizing section 21 can be provided with various image processing / editing functions. For example, a smoothing processing function is provided to smooth the image of the non-object area including the pseudo non-object block. By performing processing, for example, smoothing processing using a 3 × 3 smoothing matrix, and correcting data degraded by block distortion, a non-target object area close to the original image can be reproduced. In this case, the synthesis is performed after performing the smoothing process on the non-target object region. This is because if smoothing is to be performed after synthesis, a process of re-identifying the object block and the non-object block again and smoothing only the latter becomes necessary.

The image restored by the synthesizing unit 21 in this manner is displayed on a display unit 22 such as a color display and stored in a restored image memory unit 23 for image recording or image editing processing. It has become so. (H) Modifications Various modifications are possible in the present embodiment.

(H-1) First, the method of creating contour data described in (A) is different from that of the present embodiment in that an operator creates a desired contour by operating a known contour creating device. Alternatively, contour data may be automatically obtained by image processing using a differential operator or the like. In particular, in an image in which the object is in focus and the background part is blurred, the contour part has a large difference in shading compared to other parts, and the differential value of the pixel block shows a characteristic value. Therefore, it is also possible to obtain the differential value of the image data value for each pixel block, extract the outline component based on this value, and automatically create the outline data.

(H-2) Further, in this embodiment, the compressed data is temporarily stored in the compressed image storage unit 12 in the image data compression device 100, and this data is stored in the compressed image storage unit 13 of the image data decompression device 200. After the transfer, the image is decompressed. However, each of the image data compression device 100 and the image data decompression device 200 is provided with a data transmission unit and a data reception unit so that the compressed data is directly transmitted and received through the communication line. Alternatively, the function of the storage unit and the function of the communication line may be provided side by side and switched as appropriate.

(H-3) Further, the image data compressed by the encoding processing of each of the outline block group, the object block group, the non-object block group, and the pseudo non-object block group is written into each block group. The compressed image is stored in a different memory unit in the compressed image storage unit 12, but this has the advantage that the circuit configuration in the image data decompression device is simplified, and the compressed image of each block group depends on the type of the original image 40. Because the amount of data fluctuates,
Each memory section must have a certain amount of memory capacity, which requires an extra memory capacity.

Therefore, in a modified embodiment of the image data compression apparatus and the image data decompression apparatus described below, the compressed image data is collectively stored in the same memory unit, and when the image is decompressed, these compressed image data are stored in each block. The decoding process is performed while classifying the data into groups.

That is, in the configuration of the image data compression apparatus 101 shown in FIG. 5, the first, second, third and fourth
The image data of the outline block group, the object block group, the non-object block group, and the pseudo non-object block group compressed by the respective compression units 6, 7, 8, and And the same memory unit 24b of the compressed image storage unit 24 with the position data (x, y)
Is stored in

However, when the fourth non-target block is compressed by the fourth compression section 9, an identification symbol for distinguishing the block from the pixel blocks of the outline block group is attached to the compressed image data of the pixel block. Keep it.

On the other hand, the object pixel table is compressed by MR encoding in the fifth compression unit 10 and stored in the memory unit 24c of the compressed image storage unit 24. The object block table is stored in the sixth compression unit 10. The compressed image is compressed by the MR unit 25 and stored in the memory unit 24 a of the compressed image storage unit 24. Other operations are the same as those of the image data compression apparatus 100 of FIG.

FIG. 6 shows the image data compression apparatus 101 of FIG.
FIG. 2 is a block diagram showing a configuration of an image data decompression device 201 for decompressing image data compressed according to the first embodiment.

The compressed image storage unit 26 is the same as the compressed image storage unit 24 of the image data compression device 101. If both devices are integrated or the installation locations are close,
The compressed image storage unit 24 may be shared.

First, the compressed object block table is called from the memory section 26a of the compressed image storage section 26, decompressed by the sixth decompression section 27 and subjected to decompression processing corresponding to the sixth encoding to restore the memory. Stored in the unit 28. The pixel block determining unit 29 is a memory unit 26 of the compressed image storing unit 26.
b, the compressed image data is called for each pixel block,
The data CT (x,
Based on y), it is determined to which block group the pixel block belongs.

That is, when the value of the data CT (x, y) of the object block table corresponding to the read pixel block is “1”, the pixel block is a contour block or a pseudo non-object block. Since there is, this is sent to the switching unit 30. The switching unit 30 determines the presence or absence of an identification symbol attached to the image data of the pseudo non-object block in the third compression unit 16, and if there is the identification symbol, determines that the block is a pseudo non-object block and determines whether it is a pseudo non-object block. If there is no identification symbol, the block is determined to be a contour block and sent to the first decompression unit 14.

If the value of the data CT (x, y) in the object table corresponding to the read pixel block is "2", it can be determined that the object block is the object block. It is sent to the extension unit 15.

If the value of the data CT (x, y) of the object table corresponding to the read pixel block is "0", it can be determined that the block is a non-object block, and the compressed image data is converted to a third decompression unit. Send to 16.

It should be noted that when the decoding processes for the given pixel blocks are completed, the decompression units 14, 15, 16, and 17 feed back a signal to that effect to the pixel block discrimination unit 29. It has become
Upon receiving this end signal, the pixel block determining unit 29 sends the next pixel block to the corresponding decompressing unit.

The following operations of creating a modified contour block and synthesizing a restored image are performed by the image data restoring apparatus 20 shown in FIG.
Since it is the same as the case of 0, the description is omitted.

In this modified embodiment, the object block table is compressed by the sixth compressing section 25 and separately stored in the memory section 24 so that the compressed image data can be identified.
a, and this is decompressed by the sixth decompression unit 27 of the image data decompression device 201, and the pixel block discrimination unit 29 discriminates the pixel block based on the reconstructed object block table. However, it is not always necessary to compress and save the object block table. The CT (x, y) value of the object table is “3” of “0”, “1”, “2”.
Since this is value data, this data is
If information is input to 2 bits of the xy compressed image data as information, when the pixel block
Each pixel block group can be determined with reference to the identification data. According to such a method, the sixth compression unit 25 in the image data compression device 101 and the sixth decompression unit 27 and the memory unit 28 in the image data decompression device 201
And the like can be omitted, and the circuit configuration is simplified.

(H-4) A television monitor display section is provided in the image data compression apparatus 100, and a variable circuit is provided so that the compression ratio of each compression section can be freely set. It is also possible to achieve a fine balance between the image and the background image.

In this case, when the image data compression device 100 and the image data decompression device 200 are integrated,
The display unit 23 of the image data restoration device 200 can be used as the television monitor display unit.

When the compression ratios ra and rb or the compression ratios rc and rd in the respective compression units of the image data compression device 100 are set to the same compression ratio, the first compression unit 6 is used. And the second compression unit 7 or the third compression unit 8
And the fourth compression unit 9 can be shared by one compression unit, so that the circuit configuration can be simplified. Correspondingly, the first decompression unit 14 in the image data restoration device 200
The second and third extension units 15 and the third and fourth extension units 16 and 17 can also be shared by one extension unit, so that the circuit configuration can be similarly simplified.

(H-5) Which side of the contour of the original image is to be regarded as the target area is determined according to the situation. For example, the target block table of the image data compression apparatus 100 is created. A changeover switch is provided in part 3,
The data CT (x,
If the values of “0” and “2” in y) can be replaced, the object block group and the non-object block group can be easily converted, and the operability is greatly improved.

(H-6) In the present embodiment, each pixel block of the non-object block group is subjected to high compression processing and is stored. However, since the non-object area usually has many insignificant images, In some cases, only the average value of those block groups may be stored.

(H-7) In the image processing of the present invention, the number of target colors is not limited.
Not only monochrome, but also four colors (Y,
(M, C, K), and image processing formed by a color system such as RGB and Lab.

The processing may be performed for each color space or for each dye block of an image. After performing color conversion such as RGB-YIQ conversion used in the NTSC system on image data, The image processing according to the present invention may be performed.

[0139]

According to the image encoding method of the present invention,
For the object block group, a low compression rate encoding process is performed.
Since the encoding process with a high compression rate is performed for the non-object block group and the two processings of the low compression rate and the high compression rate are performed for the outline block group, the image quality of the important image area is maintained sharply. In addition, it is possible to prevent a density difference in the vicinity of an outline block of decompressed image data caused by a difference in encoding processing method.

The image decoding method according to the present invention is suitable for decoding image data encoded by the above-described image encoding method, and the image data belonging to the outline block group is Image data obtained by decoding two encoded data with a low compression rate and a high compression rate is obtained by dividing and synthesizing the image data by the outline of the image. An image with no difference can be obtained.

[Brief description of the drawings]

FIG. 1 is a block diagram showing an example of an image data compression device for implementing an image encoding processing method according to the present invention.

FIG. 2 is a block diagram showing an example of an image data restoring device for performing an image decoding processing method according to the present invention.

FIG. 3 is a flowchart illustrating an operation of the image data compression device of FIG. 1;

FIG. 4 is a flowchart illustrating an operation of a decompression process of compressed data in the image data restoration device of FIG. 2;

FIG. 5 is a block diagram showing another embodiment of the image data compression apparatus for performing the image encoding processing method according to the present invention.

FIG. 6 is a block diagram showing another embodiment of the image data restoring apparatus for performing the image decoding processing method according to the present invention.

FIG. 7 is a diagram illustrating an example of an original image and contour data created based on the image.

FIG. 8A is a diagram showing a pixel block of an image, and FIG. 8B is a diagram showing a state of a pixel array in the pixel block.

FIG. 9 is a diagram for explaining a procedure for creating an object block table.

FIG. 10 is a flowchart showing a procedure for setting a position corresponding to contour data in a code table to “1”.

11A is a diagram showing a data array in the contour data, and FIG. 11B is a diagram showing a data array in the object pixel table created based on the contour data.

12A is a diagram showing a state of arrangement of image data in a contour block, and FIG. 12B is a diagram showing a state of image data arrangement of a pseudo non-object table obtained by correcting data of the contour block.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Image memory part 2 Pixel block division part 3 Object block table creation part 4 Object pixel table creation part 5 Pixel block discrimination part 6 First compression part 7 Second compression part 8 Third compression part 9 Fourth Compression unit 10 Fifth compression unit 11 Pseudo non-object block creation unit 12, 13 Compressed image storage unit 14 First decompression unit 15 Second decompression unit 16 Third decompression unit 17 Fourth decompression unit 18 Fifth Expansion unit 20 modified contour block creation unit 21 synthesis unit 22 restored image memory unit 23 display unit

Claims (2)

(57) [Claims] 1. A method of encoding the pixel data to compress image data, comprising: (a) setting a contour of a predetermined object with respect to an image to form contour data; Dividing the image data into a plurality of pixel blocks; (c) based on the contour data, dividing the pixel blocks into contour blocks including the contour data; and excluding the contour data, and the object image. And (d) pixel blocks belonging to the contour block group, and a step of classifying the object block group into a non-object block group obtained by removing the contour block group and the target block group from all the pixel blocks. Are classified into object image data including the object image and non-object image data not including the object image based on the contour data, and Obtaining image data of a pseudo non-object block group by replacing with image data; and (e) performing a first encoding process with a relatively low compression ratio for a pixel block belonging to the object block group, The pixel blocks belonging to the object block group are subjected to a second encoding process with a relatively high compression ratio, and the pixel blocks belonging to the contour block group are subjected to a third encoding process with a relatively low compression ratio. And performing a fourth encoding process with a relatively high compression ratio for each of the pixel blocks belonging to the object block group to obtain encoded pixel data for the plurality of pixel blocks. Image encoding processing method. 2. A method of decoding image data encoded by the method of claim 1, wherein: (a) the image data encoded in the first encoding process is the first encoding process. Performing first decoding processing corresponding to the first decoding data, and performing second decoding processing corresponding to the second coding processing on the image data encoded in the second encoding processing. By performing the third decoding process corresponding to the third encoding process on the image data encoded by the third encoding process, the third decoded data is obtained by the fourth encoding process. Performing a fourth decoding process corresponding to the fourth encoding process on the image data encoded by the encoding process to obtain fourth decoded data, respectively, and (b) a third process based on the contour data. Decrypted data and fourth recovery Dividing the data into an object image region and a non-object image region, respectively, and generating modified decoded data by combining the third decoded data of the object image region and the fourth decoded data of the non-object image region And (c) combining the modified decoded data, the first decoded data, and the second decoded data to obtain decompressed image data of the image data.