JPH06237385A - Image coding device - Google Patents

Abstract

(57) [Abstract] [Object] The present invention relates to an image encoding device, and by filling the leftmost or lowermost pixel column of an image with the extra pixel column of the block at the rightmost or lowermost position of the image, The purpose is to maintain the continuity in blocks and improve the restoration efficiency at the time of decoding. In an image coding apparatus of an orthogonal transform system that divides an image into blocks composed of a plurality of pixels and encodes the image, when dividing the image into blocks, an image right edge P _RE is a pair of rightmost edge blocks B _RE . when located between the vertical boundary line, the pixel column X _RE image rightmost in the rightmost block BR _E, embedded in the pixel column X _RR of surplus from the image right end edge P _RE rightmost block B _RE on the right , Encode.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image coding apparatus, and more particularly, to an orthogonal transform type image coding apparatus which divides an image into blocks made up of a plurality of pixels for coding.

[0002]

2. Description of the Related Art Conventionally, an image coding apparatus for coding a multivalued image such as a color image adopts an orthogonal transform method (Hadamard transform, DCT (discrete cosine transform), etc.) as its coding method. Is the mainstream. In this orthogonal transformation method, an image is divided into blocks (for example, a rectangular block consisting of 64 pixels of 8 horizontal pixels and 8 vertical pixels) and encoded. Here, as a matter of course, depending on the image, an image having a size that is not divisible by the block size may be handled. At this time, conventionally, white (data 0) is filled in the remaining pixels in the right end block or the lower end block for encoding.

Specifically, as shown in FIG. 11, the image P
If the right edge P _RE of the right edge block _RE is located in the middle of the rightmost edge block B _RE , the extra pixel row X _RR on the right side of the right edge P _RE is white-padded and encoded. Also, the bottom edge P _{BE of the} image
Is located in the middle of the bottom block B _BE , the bottom edge P
The extra pixel row X _BR below _BE was similarly white-padded and encoded.

[0004]

However, in such a conventional image coding apparatus, a process of white-filling the pixel rows X _RR and X _BR left over from the right edge P _RE or the lower edge P _BE of the image P is processed. Was going on. This process creates a discontinuity in the block (that is, a harmonic component).
Therefore, the orthogonal transform method, which is characterized in that the harmonic components are removed and compression encoding is performed, has a problem that the restoration property is impaired.

That is, as can be seen from the comparison between the white-filled block in FIG. 11 and the coded / decoded block,
It can be seen that the image portion of the block after encoding / decoding becomes whitish as a whole as compared with that at the time of white-filling, and noise is generated at the image edge portion. As described above, the restoration efficiency is poor at the right edge and the lower edge of an image that is not divisible by a block, and as a result, an image that is difficult to see is to be decoded.

Therefore, according to the present invention, the continuity in the block is maintained and the decoding is performed by packing the rightmost edge or the lowest edge pixel row of the image in the remaining pixel rows of the block at the rightmost edge or the lowest edge position of the image. The task is to improve the restoration efficiency in time.

[0007]

The invention according to claim 1 is
In order to solve the above-mentioned problems, in an image encoding device of an orthogonal transform method that divides an image into blocks composed of a plurality of pixels and encodes the image, when dividing the image into blocks, the right edge of the image is the rightmost block. When located between a pair of vertical boundary lines, a pixel row embedding unit for embedding the rightmost pixel row of the image in the rightmost block in a pixel row on the right side of the right edge of the image of the rightmost block is provided. Is characterized by.

According to a second aspect of the present invention, in order to solve the above-mentioned problems, an image is divided into blocks in an orthogonal transform type image coding apparatus which divides an image into blocks each having a plurality of pixels for coding. In this case, when the lower edge of the image is located between the pair of vertical boundary lines of the lowermost block, the pixel row at the lowermost edge of the image in the lowermost block is set to the pixel row below the image lower edge of the lowermost block. It is characterized in that a pixel column embedding means for embedding in the pixel is provided.

According to a third aspect of the invention, in order to solve the above-mentioned problems, in addition to the structure of the first aspect, the image is divided into blocks and orthogonally transformed, and then the rightmost end of the blocks forming the image. The remaining blocks except the block are quantized using a predetermined quantization table, and the rightmost block is quantized with a larger quantization step size of the harmonic component in the horizontal direction compared to the above-mentioned predetermined quantization table. It is characterized in that a quantizing means for quantizing using a table is provided.

In order to solve the above-mentioned problems, the invention according to claim 4 is, in addition to the structure of claim 2, divided the image into blocks and orthogonally transforming them, and then, at the lowest end of the blocks constituting the image. Each of the remaining blocks except the block is quantized using a predetermined quantization table, and the bottommost block is quantized with a larger quantization step size of the harmonic component in the vertical direction compared to the above-mentioned predetermined quantization table. It is characterized in that a quantizing means for quantizing using a table is provided.

In order to solve the above-mentioned problems, the invention according to claim 5 is set according to the number of columns of the remaining pixel columns when the quantizing means quantizes, in addition to the configuration of claim 3. It is characterized in that a quantization table having a quantization step size is used. In order to solve the above-mentioned problems, the invention according to claim 6 is, in addition to the configuration of claim 4, a quantization step that is set according to the number of columns of the remaining pixel columns when the quantization means quantizes. It is characterized in that a quantization table having a size is used.

[0012]

According to the invention of claim 1, when the image is divided into blocks and the right edge of the image is located between a pair of vertical boundary lines of the rightmost block, the rightmost pixel of the image in the rightmost block. The columns are embedded in the extra pixel columns on the right side of the right edge of the image of the rightmost block. As a result, it is possible to prevent the image noise and the white tone of the image that occur when the conventional whitening is performed.

According to the second aspect of the present invention, when the image is divided into blocks, if the lower edge of the image is located between a pair of vertical boundary lines of the lowermost block, the pixel at the lowermost edge of the image in the lowermost block. A column is embedded in a pixel column below the lower edge of the image of the bottom block. As a result, it is possible to prevent the image noise and the white tone of the image that occur when the conventional whitening is performed.

According to a third aspect of the invention, in addition to the operation of the first aspect, after the image is divided into blocks and orthogonally transformed, each of the remaining blocks of the blocks forming the image except the rightmost block. The block is quantized using a predetermined quantization table, and the rightmost block is quantized using a quantization table in which the quantization step size of the harmonic component in the horizontal direction is larger than that in the predetermined quantization table. Be converted. As a result, generation of zero run length is promoted, and more efficient encoding is possible.

According to the invention of claim 4, in addition to the operation of claim 2, after the image is divided into blocks and orthogonally transformed, each of the remaining blocks of the blocks constituting the image except the bottommost block. The block is quantized using a predetermined quantization table, and the bottom block is quantized using a quantization table in which the quantization step size of the harmonic component in the vertical direction is larger than that in the predetermined quantization table. Be converted. As a result, generation of zero run length is promoted, and more efficient encoding is possible.

According to the invention of claim 5, in addition to the operation of claim 3, a quantization table having a quantization step size set according to the number of the remaining pixel rows is used when performing quantization. . As a result, more efficient encoding becomes possible. According to the invention of claim 6, in addition to the effect of claim 4, when quantizing, a quantization table having a quantization step size set according to the number of the remaining pixel columns is used. As a result, more efficient encoding becomes possible.

[0017]

DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described below with reference to the drawings. FIG. 1 is a diagram showing an embodiment of an image coding apparatus according to the present invention. First, the configuration will be described. In FIG. 1, an image encoding device 1 is a high-efficiency image encoding device using block division, and includes a block division circuit 12, an orthogonal transformer 13,
The quantizer 14 and the variable length coding unit 15 are provided.
Reference numeral 11 is an image signal input terminal, and 16 is a high efficiency code output terminal.

In recent years, orthogonal compression methods such as DCT (discrete cosine transform) have become the mainstream in compression encoding methods for multi-valued images such as color images (JPEG: Joint).
Photographic Expert Group
p, MPEG: Moving Picture Codi
ng Experts Group, H.M. 261 CC
In this embodiment, the ITT recommendation) uses this DCT orthogonal transform method.

First, the basic principle of the DCT will be described.
DCT is one of the transformations called orthogonal transformation, and transforms an image signal from the spatial domain to the frequency domain. In image coding by orthogonal transformation, the original image is usually divided into block dividing circuits 12
Is divided into blocks of N × N pixels, and processing is performed for each block. The orthogonal transformer 13 obtains N × N two-dimensional frequency components (transform coefficients) the same as the number of pixels in the original block. However, as a result of the zigzag scan shown in FIG. 2, the first coefficient indicates a DC component. Represents and continues (N × N
-1) will represent the AC component. Further, as shown in FIG. 2, the coefficient of the high frequency component in the horizontal direction is expressed as it goes to the right, and the coefficient of the high frequency component in the vertical direction is expressed as it goes down. N
The × N DCT coefficients are quantized by the quantizer 14 using a quantization table, for example, as shown in FIG. 4A, which defines a quantization step size for each coefficient. further,
The quantized DCT coefficient is variable length coded by the variable length coding unit 15. A code by orthogonal transformation such as DCT is a lossy coding method that does not completely return to the original image because it includes quantization in this way.

The DCT transformation and the inverse transformation of the 8 × 8 block are expressed by the following equations (1) and (2), respectively.

[0021]

[Equation 1]

The advantage of converting an image signal into the frequency domain in image encoding is due to the following properties of the image signal. That is, since the image signal generally has a very strong spatial correlation, most of the signal power is concentrated in the low frequency part in the frequency domain, and many signals close to zero appear in the high frequency part. That is, the weights of the coefficients are concentrated in the upper left portion of FIG. Therefore, by coding each frequency component in the frequency domain with bit allocation according to the degree of power concentration, a highly efficient image signal can be coded. Specifically, the quantization precision of the transform coefficient is finer in the low-frequency part and coarser in the high-frequency part (see FIG. 4A), and the zero-run-length coefficients are collectively set to zero as a result of the quantization. It is represented by. The DCT transform is equivalent to expanding an original image into a linear combination of base images composed of cosine waves having different frequencies (see the above equations (1) and (2)).

As described above, in orthogonal transform such as DCT, it is indispensable to divide the image into blocks of a predetermined size. However, depending on the image, it may be necessary to handle vertical and horizontal sizes that are not multiples of the block size. In this case, conventionally, as described above, white was embedded in the remaining pixels of the block at the right end or the lower end, but in the present embodiment, the encoding process is performed as follows.

Here, a case will be described in which the image coding device 1 shown in FIG. 1 is an embodiment of the image coding device according to the inventions of claims 1, 3 and 5. As described above, the image encoding device 1 of FIG. 1 is an orthogonal transform type image encoding device that divides an image into blocks each including a plurality of pixels for encoding. In the present embodiment, when the block division circuit 2 constitutes a pixel column embedding unit and divides an image into blocks, if the right edge of the image is located between the pair of vertical boundary lines of the rightmost block, The rightmost pixel row of the image in the rightmost block is embedded in the extra pixel row on the right side of the right edge of the image of the rightmost block.

That is, as shown in FIG. 3, when the image P is divided by the width of the block B from the left end, in the right end block B _RE , an extra pixel row X in which the image cannot be embedded is formed.
_RR will occur. In such a case, in the present embodiment, as shown in the central portion of FIG. 3, the right edge P _RE of the image P is
The rightmost pixel row X _RE inside the image is copied as it is to each remaining pixel row X _RR on the right side of the image right edge P _RE to generate a temporary block. Then, the pixel embedding processing as described above is performed to perform encoding and decoding. As a result, the continuity in the block can be maintained, the DC component can be held, the noise of the image edge can be suppressed, and the tone in the block can be held. Therefore, the restoration efficiency can be improved.

On the other hand, the quantizer 14 constitutes a quantizing means, divides the image P into blocks and orthogonally transforms them.
Of the blocks B constituting the image P, the remaining blocks B other than the rightmost block BR _E are quantized using a predetermined quantization table, and the rightmost block B _RE is _converted into the predetermined quantization table. In comparison, quantization is performed using a quantization table in which the quantization step size of the harmonic component in the horizontal direction is large.

These quantization tables are shown in FIG. 4, for example.
It is shown as (a) and (b). FIG. 4A shows a standard quantization table T used when quantizing the remaining blocks B excluding the rightmost block B _RE , and FIG. 4B is used for quantizing the rightmost block B _RE. The quantization table T ₀ is shown in the table, and the shaded area in each table represents a region having a relatively large step size. Then, as can be seen from FIG. 4A, the step sizes in the table T used when quantizing the remaining blocks B excluding the rightmost block B _RE are as follows: The values are set so that they are almost symmetrical with respect to the diagonal line connecting the bottom. On the other hand, in the quantization table T ₀ used for the quantization of the rightmost block, FIG.
As can be seen from (b), the quantization step size is set so that the diagonally shaded area above the diagonal line connecting the upper left and lower right in the figure is larger than the lower side area.

As a result of the zigzag scanning shown in FIG. 2, a normal block, that is, a block in which no extra pixel row is generated, should have the same spatial frequency distribution in both vertical and horizontal directions, and therefore is shown in FIG. 4 (a). The standard quantization table T has substantially the same quantization step configuration in both vertical and horizontal directions. Then, when the rightmost pixel row of the image is embedded in the remaining pixel row of the rightmost block as in this embodiment, the image becomes flat in the horizontal direction as shown in FIG. This means that the rightmost block has few horizontal harmonic components. Therefore, when performing the quantization of the orthogonal transformation coefficient of the rightmost block after performing the orthogonal transformation, a quantization table having a large quantization step size in the horizontal direction as shown in FIG. 4B is used. , The generation of zero run length is promoted, and the variable length coding of the variable length coding unit 5 enables efficient coding.

Further, preferably, when the quantizer 14 quantizes, a quantization table of a quantization step size set according to the number of columns of the remaining pixel column X _RR , for example, FIG. Quantization table T shown in b) and (c)
Try to use ₁ , T ₂ and T ₃ . Note that FIG.
4 (a), (b) and (c) are diagrams merely showing the distinction between the shaded area and the non-shaded area described in FIGS. 4 (a) and 4 (b) without displaying each quantization step size. Is. Here, if the number of remaining pixel columns in the rightmost block is 1 to 3 columns, the quantization table T ₁ is used rather than the quantization table T ₂ because the restoration efficiency is improved. In the embodiment, the quantization table is selected by the flow shown in FIG. In the flow example of FIG. 6,
When the number of remaining pixel columns at the right end of the 8 × 8 pixel block is 3 or less, the normal quantization table T ₁ of FIG. 5A is used.
When 4 pixels or more and 6 pixels or less are used,
The quantization table T ₂ in (b) is used, and the quantization table T ₃ in FIG. 5 (c) is used when there are 7 pixels. Note that H and mod in FIG. 6 mean H: horizontal size (number of pixels) of image, mod: modulo (modulo divided), and Hmod8 is an extra pixel at the right end of the 8 × 8 pixel block. Shows the number of columns. Further, S _{1 to} S ₆ in the figure represent each step of the flowchart.

As described above, in the present embodiment, by embedding the pixel row at the right end of the image in the remaining pixel row of the right end block, it is possible to surely prevent the noise and the deterioration of the basic tone due to whitening as in the conventional case. You can Also, since the quantization table is used so that the orthogonal transform coefficient of the rightmost block is roughly quantized in the horizontal direction, it is possible to promote the generation of zero run length and perform more efficient encoding. it can.

Further, according to the number of extra pixel columns in the rightmost block, that is, if the number of extra pixel columns is small, the normal quantization table is displayed. Since the encoding table is used, more efficient encoding can be performed. Next, the image coding apparatus 1 shown in FIG.
Description will be made regarding the case of an embodiment of the image coding device according to the invention described in 4 and 6.

As described above, the image coding apparatus 1 of FIG. 1 is an orthogonal transform type image coding apparatus that divides an image into blocks made up of a plurality of pixels for coding. And
In the present embodiment, when the block division circuit 2 constitutes a pixel column embedding means and divides an image into blocks, if the lower edge of the image is located between a pair of vertical boundary lines of the lowermost block, the lowermost edge is set. The pixel row at the bottom of the image in the block is embedded in the extra pixel row below the bottom edge of the image of the bottom block.

That is, as shown in FIG. 7, when division is performed from the upper end of the image P by the width of the block B, in the lower end block B _BE , an extra pixel row X in which the image cannot be embedded is formed.
_BR will occur. In such a case, in the present embodiment, as shown in FIG. 7, each of the pixel lines remaining beneath side lowermost pixel column X _BE image inside the image lower edge P _BE lower edge P _BE of the image P Copy it to X _BR as it is to generate a temporary block. Then, the pixel embedding processing as described above is performed to perform encoding and decoding. As a result, the continuity in the block can be maintained, the DC component can be held, the noise of the image edge can be suppressed, and the tone in the block can be held. Therefore, the restoration efficiency can be improved.

On the other hand, the quantizer 14 constitutes a quantizing means, divides the image P into blocks and orthogonally transforms them.
Compare each of the remaining blocks B, quantized using a predetermined quantization table, the lowermost end block B _BE, the predetermined quantization tables, except for the lowermost block BB _E of the blocks B constituting the image P Then, quantization is performed using a quantization table in which the quantization step size of the harmonic component in the vertical direction is large. These quantization tables are shown in FIG.
It is shown as (a) and (b). FIG. 8A shows a standard quantization table T used when quantizing each of the remaining blocks B excluding the bottom block B _BE , and FIG. 8B is used for the quantization of the bottom block B _BE. The quantization table T ₁₀ is shown, and the shaded portion in each table represents a region having a relatively large step size. Then, as can be seen from FIG. 8A, the step sizes in the table T used when quantizing the remaining blocks B excluding the bottommost block B _BE are the upper left and the right in the figure of the table. The values are set so that they are almost symmetrical with respect to the diagonal line connecting the bottom. On the other hand, in the quantization table T ₁₀ used for the quantization of the bottommost block, FIG.
As can be seen from (b), the quantization step size is set so that the shaded area below the diagonal connecting the upper left and lower right in the figure is larger than the upper area.

A normal block, that is, a block in which no extra pixel row is generated, should have the same spatial frequency distribution in the vertical and horizontal directions as a result of the zigzag scanning shown in FIG. 2, so that it is shown in FIG. 8A. The standard quantization table T has substantially the same quantization step configuration in both vertical and horizontal directions. Then, when the bottommost pixel row of the image is embedded in the remaining pixel row of the bottommost block as in this embodiment, the image becomes flat in the vertical direction as shown in FIG. This means that the lowermost block has few vertical harmonic components. Therefore, when performing the quantization of the orthogonal transformation coefficient of the bottommost block after performing the orthogonal transformation, by using a quantization table having a large quantization step size in the vertical direction as shown in FIG. 8B. , The generation of zero run length is promoted, and the variable length coding of the variable length coding unit 5 enables efficient coding.

Further, preferably, when the quantizer 14 quantizes, a quantization table of a quantization step size set according to the number of the remaining pixel rows X _BR , for example, FIG. The quantization tables T ₁₁ , T ₁₂ and T ₁₃ shown in b) and (c) are used. Note that FIG.
(A), (b), and (c) are diagrams only showing the distinction between the shaded area and the non-shaded area described in FIGS. 8A and 8B without displaying each quantization step size. Is. Here, if the remaining pixel columns in the lowermost block are the 1st to 3rd columns, the quantization table T ₁₁ is more efficient than the quantization table T ₁₂ because the restoration efficiency is higher. In the embodiment, the quantization table is selected by the flow shown in FIG. In the flow example of FIG. 10, the number of remaining pixel columns at the lower end of the 8 × 8 pixel block is 3
If the number is less than or equal to 4 and the normal quantization table T ₁₁ in FIG.
The quantization table T ₁₂ of 0 (b) is used, and when the number of pixels is 7, the quantization table T ₁₃ of FIG. 10 (c) is used. Note that V and mod in FIG. 10 respectively mean V: vertical size (number of pixels) of the image, mod: modulo (modulo divided), and Vmod8 is an extra pixel at the lower end of the 8 × 8 pixel block. Shows the number of columns. Also,
In the figure, S _{11 to} S ₁₆ represent steps of the flowchart.

As described above, in the present embodiment, by embedding the pixel row at the bottom end of the image in the remaining pixel row at the bottom end block, it is possible to reliably prevent noise and deterioration of the basic tone due to white padding as in the conventional case. You can Further, since the quantization table is used so that the orthogonal transform coefficient of the bottommost block is roughly quantized in the vertical direction, it is possible to promote the generation of zero run length and perform more efficient encoding. it can.

Further, according to the number of extra pixel columns in the bottommost block, that is, if the number of extra pixel columns is small, the normal quantization table is displayed. Since the encoding table is used, more efficient encoding can be performed.

[0039]

According to the present invention, the rightmost edge or the lowest edge pixel row of the image is embedded in the remaining pixel rows of the block at the rightmost edge or the lowest edge position of the image. Compared with the above, the continuity in the block can be maintained and the restoration efficiency at the time of decoding can be improved.

[Brief description of drawings]

FIG. 1 is a block diagram showing an embodiment of an image encoding device according to the present invention.

FIG. 2 is a diagram illustrating the relationship between the DCT coefficient and zigzag scanning.

FIG. 3 is a diagram for explaining an operation of embedding a pixel column, showing an embodiment of the image coding apparatus according to claims 1, 3 and 5;

FIG. 4 is a diagram showing a quantization table thereof.

FIG. 5 is a diagram showing a step size pattern of the quantization table.

FIG. 6 is a diagram showing the quantization table selection flow.

FIG. 7 is a diagram for explaining an operation of embedding a pixel column, showing an embodiment of the image encoding device according to claims 2, 4 and 6;

FIG. 8 is a diagram showing the quantization table.

FIG. 9 is a diagram showing a step size pattern of the quantization table.

FIG. 10 is a diagram showing the quantization table selection flow.

FIG. 11 is a diagram illustrating a pixel column embedding operation in a conventional image encoding device.

[Explanation of symbols]

 11 Input Terminal 12 Block Division Circuit 13 Orthogonal Transformer 14 Quantizer 15 Variable Length Coding Section 16 Output Terminal

Claims (6)

[Claims] 1. An image coding apparatus of an orthogonal transform system, which divides an image into blocks composed of a plurality of pixels and encodes the blocks, wherein when the image is divided into blocks, a right edge of the image is a pair of rightmost blocks. When located between the vertical boundary lines, a pixel row embedding unit is provided for embedding the rightmost pixel row of the image in the rightmost block in the extra pixel row on the right side of the right edge of the image of the rightmost block. The image coding device. 2. An image coding apparatus of an orthogonal transform system for coding an image by dividing it into blocks made up of a plurality of pixels and, when dividing the image into blocks, a lower edge of the image is a pair of lowermost blocks. When located between the vertical boundary lines, a pixel column embedding unit for embedding the pixel column at the lowermost end of the image in the lowermost block into the pixel column below the lower edge of the image of the lowermost block is provided. Image encoding device. 3. The image is divided into blocks and orthogonally transformed, and then the remaining blocks of the blocks forming the image except the rightmost block are quantized using a predetermined quantization table, and The quantizing means for quantizing the right end block by using a quantizing table having a larger quantizing step size of a harmonic component in the horizontal direction as compared with the predetermined quantizing table is provided. The image encoding device described. 4. The image is divided into blocks and orthogonally transformed, and then the remaining blocks of the blocks constituting the image except the lowermost block are quantized using a predetermined quantization table, The quantizing means for quantizing the lower end block by using a quantizing table having a larger quantizing step size of a harmonic component in the vertical direction as compared with the predetermined quantizing table is provided. The image encoding device described. 5. A quantization table having a quantization step size set according to the number of the extra pixel columns is used when the quantizing means quantizes. Image encoding device. 6. A quantization table having a quantization step size set according to the number of the remaining pixel rows is used when the quantization means quantizes. Image encoding device.