JP3466162B2 - Image decoding apparatus and image decoding method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an image encoding device, an image decoding device, an image encoding method, and an image decoding method which can perform encoding and decoding more efficiently than when conventional binary image encoding and decoding technologies are used by predicting encoded pixels from a previously obtained binary image having high correlation and encoding their differences, and providing a medium in which programs for making a compute execute those operations are recorded. SOLUTION: This device is provided with a block dividing means 1(101) which inputs an object binary image to be encoded and divides the object binary image into blocks containing pixels to obtain an object block, a block dividing means 2(102) which divides the previously obtained reference binary image into blocks containing pixels to obtain a reference block, an exclusive OR block constituting means (103) which scans the said object block and reference block in order to detect exclusive OR between both pixel values and constitute an exclusive OR block, and an exclusive OR encoding means (104) which generates a code sequence of the exclusive OR and outputs it as encoded data.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image decoding apparatus and an image decoding method which can be used for transmitting and storing images, especially binary moving images.

[0002]

2. Description of the Related Art Conventionally, when synthesizing an image, in addition to the brightness of an object, there is a case where information called an alpha value indicating an occupied area of the object and a transparency is added. The alpha value is determined for each pixel, and 1 means opaque or occupied, and 0 means completely transparent or unoccupied. That is, when the image of a certain object is embedded in the background image, the alpha value is required. Hereinafter, an image having only this alpha value is called an alpha plane.

In the case of clouds, frosted glass, etc., the alpha value may be represented by an intermediate value of [0, 1],
In some cases, a binary value of {0, 1} is sufficient.

CCITT, which has been used in conventional facsimiles, is used for encoding general binary alpha planes.
Binary image coding technology MR, MMR
Encoding or encoding standardized by JBIG can be used. These are collectively called still binary image coding.
In still binary image coding, it is possible to efficiently code by predicting a lower pixel from a higher pixel in the scanning direction and entropy coding the difference.

[0005]

In a binary moving image such as two continuous alpha planes of a moving image, correlation can be used between consecutive frames. That is,
The correlation obtained in advance is higher than that of predicting the lower pixel from the upper pixel in the scanning direction and encoding the difference.
It is more efficient to predict the coded pixel from the value image and code the difference.

However, in the conventional still binary image coding,
Even if each of the encoding device and the decoding device obtains a binary image having a very high correlation with the binary image to be encoded or decoded, only the upper and lower correlations in the scanning direction are shown. However, it has a problem that a large amount of code is required.

The present invention, in consideration of the above problems of the prior art, provides an image decoding apparatus and an image decoding method capable of performing decoding more efficiently than using the conventional binary image decoding technique. The purpose is to provide.

[0008]

The present invention according to claim 1 decodes coded data obtained by coding a binary image showing the transparency information of each frame of a moving image for each block, In an image decoding device for obtaining a binary image of a block , refer to a frame decoded before a target frame having the target block as a decoding target .
A reference block generation means for obtaining a reference block is a binary image, the ginseng corresponding to the target pixel in the previous SL target block
Statistical pixel selection means for selecting a statistical model from a plurality of statistical models based on the states of reference pixels in the illumination block and the states of pixels surrounding the reference pixel, and the code based on the selected statistical model. An image decoding device comprising: an arithmetic decoding means for decoding the encoded data to obtain the binary image of the target block.

The present invention according to claim 2 is characterized in that the peripheral pixel of the reference pixel is a peripheral pixel located within one pixel range around the reference pixel. It is a decoding device.

The present invention according to claim 3 is the image decoding apparatus according to claim 1, characterized in that the peripheral pixels of the reference pixel are four pixels which are located above and below and to the left and right of the reference pixel. Is.

According to the present invention of claim 4, the statistical model selecting means is based on the states of the peripheral pixels of the target pixel in addition to the states of the reference pixels and the peripheral pixels of the reference pixels. The image decoding apparatus according to any one of claims 1 to 3, wherein the image decoding apparatus is a statistical model selection unit that selects a statistical model from a plurality of statistical models.

The present invention according to claim 5 is characterized in that the peripheral pixels of the target pixel are located above, to the left, to the upper left and to the upper right of the target pixel. It is a device.

According to the present invention of claim 6 , each of the moving images is
In an image decoding method of decoding a coded data obtained by coding a binary image indicating frame transparency information for each block to obtain a binary image of the target block, the target block to be decoded is a reference binary image generating step of obtaining a reference binary image from the decoded frame even prior to the target frame having a reference block generating step of obtaining a reference block containing a plurality of pixels from the reference binary image, before Symbol The above corresponding to the target pixel in the target block
A statistical model selecting step of selecting a statistical model from a plurality of statistical models based on a state of reference pixels in a reference block and states of pixels surrounding the reference pixel; and the code based on the selected statistical model. And an arithmetic decoding step of decoding the encoded data to obtain a binary image of the target block.

Further, according to the present invention of claim 7, the peripheral pixel of the reference pixel is a peripheral pixel located within one pixel range around the reference pixel. This is a decoding method.

The present invention according to claim 8 is characterized in that the surrounding pixels of the reference pixel are four pixels located above, below, to the left and right of the reference pixel. Is.

Further, in the present invention according to claim 9, the statistical model selecting step is based on the states of the peripheral pixels of the target pixel, in addition to the states of the reference pixel and the peripheral pixels of the reference pixel. 9. The image decoding method according to any one of claims 6 to 8, which is a statistical model selection step of selecting a statistical model from a plurality of statistical models.

According to the present invention of claim 10, the surrounding pixels of the target pixel are above, left, upper left,
The image decoding method according to claim 9, wherein the image decoding method is located at the upper right.

[0018]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described below with reference to the drawings. (Embodiment 1) FIG. 1 is a block diagram showing a configuration of an image coding apparatus according to an embodiment of the present invention. In FIG. 1, a blocking unit 1 (101) is a unit that receives an image to be encoded as an input and divides the input image into blocks each including a plurality of pixels. Blocking means 2
(102) is a unit that divides the reference image obtained in advance into blocks each including a plurality of pixels. The exclusive OR block forming means (103) is a block forming means 1 (10).
Target block divided by 1) and blocking means 2
This is means for scanning the reference block divided by (102), calculating the exclusive OR of the pixel values thereof, and forming an exclusive OR block. Exclusive-OR encoding means (104)
Is a means for encoding the exclusive OR block and outputting the encoded data.

The operation of the image coding apparatus of the present embodiment configured as described above will be described below, and the operation of one embodiment of the image coding method of the present invention will be described at the same time.

That is, the mask moving image (19
The t + 1th frame (1903) of 01) is the target binary image, and the tth frame (1902) is the reference binary image. Hereinafter, in the figure, 1 is represented by black and 0 is represented by white. The target binary image (1903) is divided into 8 pixels × by the blocking unit 1 (101) like the target divided image (1905).
It is divided into target blocks of 8 pixels. Blocking means 1
(101) is not limited to division of 8 pixels × 8 pixels or 16 pixels × 16 pixels.

The reference binary image (1902) is divided into reference blocks of 8 pixels × 8 pixels by the blocking means 2 (102) like the reference divided image (204). In the block forming means 2 (102), 8 pixels × 8 pixels or 16 pixels
The division is not limited to pixels × 16 pixels.

The target block (2002) in FIG. 20 is one of the blocks in the target divided image (1905).
The reference block (2001) is the target block (2002)
Is a block in the reference divided image (1904) corresponding to. The exclusive OR block configuration means (103) causes the target block (2002) and the reference block (200).
1) is scanned from the upper left to the lower right, the exclusive OR of pixel values is calculated, and the exclusive OR block (2003) is configured. Exclusive OR block (200
3) is encoded by the exclusive OR encoding means (104) using a method generally called arithmetic encoding.
A brief description of arithmetic coding is given below (see "International Standards for Multimedia Coding", Chapter 3, Arithmetic Coding, Hiroshi Yasuda, Maruzen Co., Ltd.).

FIG. 21 is a diagram for explaining the principle of arithmetic coding. In arithmetic coding, a symbol string (2105)
By the symbol occurrence probability model (2104), the number line (210
The range obtained by limiting (1) and continuing (210)
2) The shortest binary point (2103)
Is output as encoded data.

FIG. 22 shows a flowchart of arithmetic coding. At 2201, arithmetic coding is started. At 2202, the range is initialized from 0 to 1. At 2203, the symbol is input. At 2204, the occurrence probability model is assigned to the current range, and the range of the probability of the input symbol is set as the new range. At 2205, if the symbol is the ending symbol,
At 2206, the range is represented by a binary point, the binary point is output, and at 2207, the arithmetic coding ends. 2205,
If the symbol is not the end symbol, the next symbol is input at 2203. However, if the number of symbols is fixed, the end symbol can be omitted.

Decoding is performed by determining the symbol sequence from the binary point. It is known that arithmetic coding has the property that the code amount of a symbol string decreases as the symbols and the occurrence probability models of the symbols match and the occurrence probabilities of the symbols are biased. . It is also known that even if the occurrence probability model is changed during encoding, it can be decoded if the way of change is known.

In the exclusive OR encoding means (104), the above arithmetic encoding and [0, 0.9) are performed for the symbol 0,
Using the occurrence probability model in which [0.9, 1.0) is the symbol 1, the code string of the exclusive OR block that is the symbol string of 0 and 1 is generated and the code data is output.

As described above, in the present embodiment, in the case of a mask moving image, the occurrence probability of the symbol 0 and the symbol 1 of the exclusive OR of the target block and the reference block is about 9: 1. It is possible to perform efficient encoding with a small code amount by combining the exclusive OR with the arithmetic encoding by using the. (Embodiment 2) FIG. 2 is a block diagram showing a configuration of an image decoding apparatus according to an embodiment of the present invention, and the configuration of the present embodiment will be described with reference to FIG. In the figure, an exclusive OR decoding means (201) is means for receiving encoded data and decoding it to obtain an exclusive OR block.
The blocking unit 2 (202) is a unit that divides a reference image obtained in advance into reference blocks composed of a plurality of pixels. The target block constructing means (203) is means for obtaining the target block from the exclusive OR block obtained by the exclusive OR decoding means (201) and the reference block obtained by the blocking means (202).

The operation of the image decoding apparatus of the present embodiment configured as described above will be described below, and the operation of one embodiment of the image decoding method of the present invention will be described at the same time.

That is, the exclusive OR decoding means (201)
Is the same as the exclusive OR encoding (104) [0, 0.
9) is a decoder of an arithmetic coding system having an occurrence probability model in which the symbol 0 is set to 9) and the symbol 1 is set to [0.9, 1.0). A symbol string is generated from the binary point which is the code data and the occurrence probability model, and the symbols are arranged in the scanning direction to form an exclusive OR block.

The blocking means 2 (202) is equivalent to the blocking means 2 (102). The target block constructing means (203) scans the exclusive OR block and the reference block, and the pixel having the exclusive OR block of 1 inverts the pixel value of the reference block to obtain the target block.

As described above, in the present embodiment, in the case of a mask moving image, the occurrence probability of the symbol 0 and the symbol 1 of the exclusive OR of the target block and the reference block is about 9: 1. It is possible to perform efficient decoding with a small code amount by combining the exclusive OR and the arithmetic coding by using the. (Embodiment 3) FIG. 3 is a block diagram showing a configuration of an image coding apparatus according to an embodiment of the present invention, and the configuration of the present embodiment will be described with reference to FIG. In the figure, a blocking unit 1 (301) is a unit that receives an image to be encoded as an input and divides the input image into blocks composed of a plurality of pixels. The motion estimation means (305)
This is a means for searching a block similar to the target block in the reference image and obtaining its motion vector. The motion compensation block formation unit 2 (302) is a unit that receives a reference image and motion information and divides the input reference image into blocks of a plurality of pixels based on the motion information. The exclusive OR block construction means (303) is a block formation means 1 (3
This is a means for scanning the target block divided by 01) and the reference block divided by the motion compensation block formation means 2 (302), calculating the exclusive OR of the pixel values thereof, and forming an exclusive OR block. The exclusive OR encoding means (304) is means for encoding the exclusive OR block and outputting the encoded data.

The operation of the image coding apparatus of the present embodiment configured as described above will be described below, and the operation of one embodiment of the image coding method of the present invention will be described at the same time.

First, the blocking means 1 (301) is equivalent to the blocking means 1 (101). A motion estimating means (305) is a motion vector to be estimated, v is the number of pixels in the target block, and u_i (i is the position of each pixel in the image.
Is 1 to m), and the pixel value at the position x in the target image is A
(X), assuming that the pixel value at the position x in the reference image is B (x), the similarity s (v) is minimized v using (Equation 1).
Is detected from a predetermined range, and v is output as a motion vector.

[0034]

[Equation 1] The motion compensation block formation means (302) shifts the block cut out from the reference image by the motion vector v, obtains the reference block, and outputs it. The exclusive OR block configuration means (303) is an exclusive OR block configuration means (103).
Is equivalent to The exclusive OR encoding means (304) is equivalent to the exclusive OR encoding means (104).

As described above, according to this embodiment, the occurrence probabilities of the symbol 0 and the symbol 1 of the exclusive OR block are calculated by using the motion estimation means and the motion compensation block formation means.
It is possible to perform efficient coding with a small code amount by performing motion compensation on a block that is significantly different from 9: 1 so as to be close to 9: 1. (Embodiment 4) FIG. 4 is a block diagram showing a configuration of an image decoding apparatus according to an embodiment of the present invention, and the configuration of the present embodiment will be described with reference to FIG. In the figure, an exclusive OR decoding means (401) is means for receiving encoded data and decoding it to obtain an exclusive OR block.
The motion compensation block formation unit 2 (402) is a unit that receives a reference image and motion information and divides the input reference image into blocks composed of a plurality of pixels based on the motion information. The target block constructing means (403) is means for obtaining a target block from the exclusive OR block obtained by the exclusive OR decoding means (401) and the reference block obtained by the motion compensation block forming means (402). is there.

The operation of the image decoding apparatus of the present embodiment configured as described above will be described below, and the operation of one embodiment of the image decoding method of the present invention will be described at the same time.

That is, exclusive OR decoding means (401)
Is equivalent to the exclusive OR decoding means (201). The motion compensation block formation means 2 (402) is equivalent to the motion compensation block formation means 2 (302). The target block composing means (403) is equivalent to the target block composing means (203).

As described above, according to this embodiment, the occurrence probabilities of the symbol 0 and the symbol 1 of the exclusive OR block are calculated by using the motion estimation means and the motion compensation block formation means.
It is possible to perform efficient decoding with a small code amount by performing motion compensation on a block that is significantly different from 9: 1 so as to be close to 9: 1. (Embodiment 5) FIG. 5 is a block diagram showing a configuration of an image coding apparatus according to an embodiment of the present invention, and the configuration of the present embodiment will be described with reference to FIG.

In the figure, blocking means 1 (50
1) is a means that receives an image to be encoded as an input and divides the input image into blocks each including a plurality of pixels. The blocking unit 2 (502) is a unit that receives a reference image as an input and divides the input reference image into blocks each including a plurality of pixels. Exclusive OR block construction means (5
03) scans the target block divided by the blocking unit 1 (501) and the reference block divided by the blocking unit 2 (502), calculates the exclusive OR of the pixel values, and calculates the exclusive OR block. Is a means of configuring. The exclusive OR encoding means (504) is means for encoding the exclusive OR block and outputting the encoded data. The reference block usage determining means (505) is means for comparing the target block with the reference block, outputting a reference block usage determining signal, and switching the subsequent processing. The target pixel encoding means (506) is means for encoding the target block and outputting the encoded data.

The operation of the image coding apparatus of the present embodiment configured as described above will be described below, and the operation of one embodiment of the image coding method of the present invention will be described at the same time.

First, the blocking means 1 (501) is equivalent to the blocking means 1 (101). The blocking means 2 (502) is equivalent to the blocking means 2 (102). The reference block use determining means (505) uses the absolute value sum (SAD) of the target block and the reference block, and if the absolute value sum is greater than or equal to a threshold value, encodes it using the target pixel encoding means (506) to obtain the absolute value. If the total sum is less than a certain threshold value, switching is performed so that the exclusive OR block forming means (503) and the exclusive OR encoding means (504) are used for encoding, and a reference block use determination signal is output. 5 is used as the threshold value. The exclusive OR block configuration means (503) is an exclusive OR block configuration means (103).
Is equivalent to The exclusive OR encoding means (504) is equivalent to the exclusive OR encoding means (104). The target pixel encoding means (506) is an exclusive OR encoding means (50
4), the input is the target block, and [0,
This is an arithmetic encoder using an occurrence probability model in which 0.5) is a symbol 0 and [0.5, 1.0) is a symbol 1.

As described above, according to the present embodiment, the reference block use determining means determines that a block in which the occurrence probabilities of symbol 0 and symbol 1 are significantly different from 9: 1 is a block having a large sum of absolute values. Thinking and changing the coding method, reduce the blocks with poor coding efficiency,
Efficient coding with a small code amount is possible. (Embodiment 6) FIG. 6 is a block diagram showing a configuration of an image decoding apparatus according to an embodiment of the present invention, and the configuration of the present embodiment will be described with reference to FIG. In the figure, an exclusive OR decoding means (601) is a means for receiving encoded data and decoding it to obtain an exclusive OR block.
The blocking means 2 (602) receives the reference image as an input,
It is a means for dividing the input reference image into reference blocks composed of a plurality of pixels. Target block configuration means (603)
Is means for obtaining a target block from the exclusive OR block obtained by the exclusive OR decoding means (601) and the reference block obtained by the blocking means (602). The reference block usage control means (604) is a means for switching the subsequent processing according to the reference block usage determination signal. The target pixel decoding means (605) is a means for decoding the encoded data to obtain a target block.

The operation of the image decoding apparatus of the present embodiment configured as described above will be described below, and the operation of one embodiment of the image decoding method of the present invention will be described at the same time.

That is, exclusive OR decoding means (601)
Is equivalent to the exclusive OR decoding means (201). The blocking means 2 (602) is the blocking means 2 (10
It is equivalent to 2).

The target block constructing means (603) is equivalent to the target block constructing means (203). The reference block use control means (604) performs exclusive OR decoding with the target block forming means (603), the blocking means 2 (602), and the subsequent block processing when the reference block is used, in accordance with the reference block use determination signal. Switching is made between the case of the encoding means (601) and the case of the target pixel decoding means (605) when the reference block is not used.

The target pixel decoding means (605) uses the occurrence probability model in which [0, 0.5) is the symbol 0 and [0.5, 1.0) is the symbol 1 as in the target pixel encoding (506). It is a decoder of an arithmetic coding system having. A symbol string is generated from the binary point which is the code data and the occurrence probability model, and the symbols are arranged in the scanning direction to form the target block.

As described above, according to the present embodiment, the reference block use determining means determines that a block in which the occurrence probabilities of symbol 0 and symbol 1 are significantly different from 9: 1 is a block having a large sum of absolute values. Thinking and changing the coding method, reduce the blocks with poor coding efficiency,
Efficient decoding with a small code amount is possible. (Embodiment 7) FIG. 7 is a block diagram showing a configuration of an image coding apparatus according to an embodiment of the present invention, and the configuration of the present embodiment will be described with reference to FIG. In the figure, a blocking unit 1 (701) is a unit that receives an image to be encoded as an input and divides the input image into blocks each including a plurality of pixels.

The blocking means 2 (702) is a means for inputting a reference image and dividing the input reference image into blocks each having a plurality of pixels. Statistical model selection means (7
03) is the position of the pixel to be encoded, the reference block,
It is a means for inputting a statistical model table to be described later, selecting a statistical model from the statistical model table (704) according to the states of surrounding pixels in the reference block of the target pixel, and outputting it to the entropy coding means (705). That is,
The statistical model selection unit 703 is a unit that selects a statistical model from a plurality of statistical models based on the states of pixels around the reference pixel in the reference block including the reference pixel corresponding to the target pixel in the target block. The entropy coding unit (705) outputs the position of the pixel to be coded to the statistical model selection unit (703), entropy-codes the target block based on the statistical model output from the statistical model selection unit (703), It is a means for outputting encoded data.

The operation of the image coding apparatus of the present embodiment configured as described above will be described below, and the operation of one embodiment of the image coding method of the present invention will be described at the same time.

First, the blocking means 1 (701) is equivalent to the blocking means 1 (101). The blocking means 2 (702) is equivalent to the blocking means 2 (102).

The statistical model selecting means (703) selects a statistical model from a plurality of statistical models and outputs it to the entropy coding means (705). As shown in FIGS. 23 to 26, the statistical model table (2301) is a table in which an index is assigned to each state of surrounding pixels in the reference block of the target pixel, and a statistical model is assigned to each index. The correspondence between each state and the index will be described with reference to FIGS. 27 and 28.

Since the surrounding state is considered, first, the reference block (2401) is extrapolated to create an extrapolated reference block. As for the creation method, if the surrounding pixel values are obtained from the reference image, they are added to create the extrapolation reference block (2402).

If not obtained, the extra pixels are added as they are to create an extrapolation reference block (2403). Similarly, an extrapolation target block is created from the target block. The state of the surrounding pixels in the reference block of the target pixel is shown in FIG.
8, the reference block (2501) is obtained by applying the reference mask (2503) and the target block (2502) is obtained by applying the target mask (2504). Reference mask (250
3) and the pixel values at the respective positions of the target mask (2504) are A, B, C, D, E, F, G, H, I, J,
Assuming K, L, and M, the index i is represented by (Equation 2). Further, in FIG. 28, reference numeral 2501a is attached to the reference pixel corresponding to the target pixel 2502a. In addition, the statistical model selection unit 703 of the present embodiment uses the extrapolation target block created above in order to take into consideration the states of the neighboring pixels of the target pixel 2502a already encoded in the target block. Similarly to the target pixel 2502
Get the states of neighboring pixels of a. As a result, a more appropriate statistical model can be selected as compared with the case where only the states of the surrounding pixels in the reference block are used. Of course, a configuration using only the states of surrounding pixels in the reference block may be used.

[0054]

[Equation 2] At this time, the statistical model of the index i of the statistical model table (2301) is selected.

As described above, the statistical model selection means (70
In 3), a statistical model is selected from the statistical model table and output to the entropy coding means (705). The entropy encoding means (705) uses an arithmetic encoder like the exclusive OR encoding means (104),
As the occurrence probability model of the arithmetic encoder, the statistical pixel (704) selected by the statistical model selecting means (703) is used to encode the target pixel.

As described above, according to the present embodiment, since the statistical model is switched and used by the statistical model selection means in the state of the surrounding pixels in the reference block of the target pixel, the efficiency of entropy coding is improved and the code is It enables efficient coding with a small amount. (Embodiment 8) FIG. 8 is a block diagram showing a configuration of an image decoding apparatus according to an embodiment of the present invention, and the configuration of the present embodiment will be described with reference to FIG.

In the figure, blocking means 2 (80
2) is a means for inputting a reference image and dividing the input reference image into blocks each including a plurality of pixels. The statistical model selection means (803) is a position of the encoding target pixel,
It is a means for inputting a reference block and a statistical model table, selecting a statistical model from the statistical model table (704) according to the states of pixels surrounding the reference block of the target pixel, and outputting it to the entropy coding means (705). The entropy decoding means (801) is means for receiving the encoded data and decoding the encoded data based on the statistical model (804) to obtain the target block.

The operation of the image decoding apparatus of the present embodiment configured as described above will be described below, and the operation of one embodiment of the image decoding method of the present invention will be described at the same time.

First, the blocking means 2 (802) is equivalent to the blocking means 2 (102). The statistical model selection means (803) is equivalent to the statistical model selection means (703).

In the entropy decoding means (801),
An arithmetic code decoder is used like the exclusive OR decoding means (201), but the statistical model (804) selected by the statistical model selection means (803) is used as the occurrence probability model of the arithmetic encoder. . Statistical model table (80
4) is equivalent to the statistical model table (704).

As described above, according to the present embodiment, since the statistical model is switched and used by the statistical model selection means depending on the state of the surrounding pixels in the reference block of the target pixel, the efficiency of entropy coding is improved and the coding is performed. A small amount of efficient decoding is possible. (Embodiment 9) FIG. 9 is a block diagram showing a configuration of an image coding apparatus according to an embodiment of the present invention, and the configuration of the present embodiment will be described with reference to FIG.

In the figure, blocking means 1 (90
1) is a means that receives an image to be encoded as an input and divides the input image into blocks each including a plurality of pixels. The motion estimation means (906) is means for searching a block similar to the target block in the reference image and obtaining the motion vector thereof. Motion compensation block forming means 2 (90
2) is a means for inputting a reference image and motion information, and dividing the input reference image into blocks composed of a plurality of pixels based on the motion information. Statistical model selection means (903)
Is the position of the pixel to be encoded, the reference block, and the statistical model table as input, selects the statistical model from the statistical model table (904) according to the states of the surrounding pixels of the reference block of the target pixel, and entropy encoding means ( 90
It is a means for outputting to 5). The entropy coding means (905) is means for entropy coding the target block based on the statistical model selected by the statistical model selection means (903) and outputting the coded data.

The operation of the image coding apparatus of the present embodiment configured as described above will be described below, and the operation of one embodiment of the image coding method of the present invention will be described at the same time.

First, the blocking means 1 (901) is equivalent to the blocking means 1 (101). The motion estimation means (906) is equivalent to the motion estimation means (305).
The motion compensation blocking means 2 (902) is equivalent to the motion compensation blocking means 2 (302).

The statistical model selecting means (903) is equivalent to the statistical model selecting means (703). The statistical model table (904) is equivalent to the statistical model table (704). The entropy coding means (905) is equivalent to the entropy coding means (705).

As described above, according to this embodiment, the motion estimation means and the motion compensation block forming means can be used to improve the accuracy of the statistical model and enable efficient encoding with a small code amount. (Embodiment 10) FIG. 10 is a block diagram showing a configuration of an image decoding apparatus according to an embodiment of the present invention, and the configuration of the present embodiment will be described with reference to FIG.

In the figure, the motion compensation block forming means 2
(1002) is a means for inputting a reference image and motion information, and dividing the input reference image into blocks composed of a plurality of pixels based on the motion information. The statistical model selection means (1003) receives the position of the encoding target pixel, the reference block, and the statistical model table as input, and selects the statistical model from the statistical model table (1004) according to the states of the surrounding pixels of the reference block of the target pixel. And outputs it to the entropy coding means (1005). The entropy decoding unit (1001) is a unit that receives the encoded data and decodes the encoded data based on the statistical model to obtain the target block.

The operation of the image decoding apparatus of the present embodiment configured as described above will be described below, and the operation of one embodiment of the image decoding method of the present invention will be described at the same time.

That is, the motion compensation block forming means 2 (100
2) is equivalent to the motion compensation blocking means 2 (402). The statistical model selection means (1003) is equivalent to the statistical model selection means (803). The entropy decoding means (1001) has an entropy decoding means (80
It is equivalent to 1). Statistical model table (1004)
Is equivalent to the statistical model table (704).

As described above, according to the present embodiment, by using the motion compensation block formation means 2, the accuracy of the statistical model can be improved and efficient decoding with a small code amount can be performed. (Embodiment 11) FIG. 11 is a block diagram showing a configuration of an image coding apparatus according to an embodiment of the present invention, and the configuration of the present embodiment will be described with reference to FIG.

In the figure, blocking means 1 (110
1) is a means that receives an image to be encoded as an input and divides the input image into blocks each including a plurality of pixels.

Blocking means 2 (1102) is a means for inputting a reference image and dividing the input reference image into blocks composed of a plurality of pixels. The statistical model selection means (1103) receives the position of the encoding target pixel, the reference block, and the statistical model table as input, and selects the statistical model from the statistical model table (1104) according to the states of the surrounding pixels of the reference block of the target pixel. And outputs it to the entropy coding means (1105). The reference block utilization determining means (1105) is a means for comparing the target block with the reference block, outputting a reference block utilization determining signal, and switching the subsequent processing. The target pixel encoding means (1106) is means for encoding the target block and outputting the encoded data. The entropy coding means (1105) is based on the statistical model (1104),
This is means for entropy-encoding the target block and outputting the encoded data. Target pixel encoding means (1106)
Is a means for encoding the target block and outputting the encoded data.

The operation of the image coding apparatus of the present embodiment configured as described above will be described below, and the operation of one embodiment of the image coding method of the present invention will be described at the same time.

First, the blocking means 1 (1101)
It is equivalent to the blocking means 1 (101). The blocking means 2 (1102) is equivalent to the blocking means 2 (102). The statistical model selection means (1103) is equivalent to the statistical model selection (703). The statistical model table (1104) is equivalent to the statistical model table (704). The entropy coding means (1105) is equivalent to the entropy coding means (705). The reference block usage determining means (1106) is equivalent to the reference block usage determining means (505). The target pixel decoding means (1107) is equivalent to the target pixel decoding means (605).

As described above, according to the present embodiment, the reference block utilization control means changes the coding method for blocks that do not match the statistical model, thereby reducing the blocks having poor coding efficiency and reducing the coding. A small amount of efficient decoding is possible. (Embodiment 12) FIG. 12 is a block diagram showing a configuration of an image decoding apparatus according to an embodiment of the present invention, and the configuration of the present embodiment will be described with reference to FIG.

In the figure, blocking means 2 (120
2) is a means for inputting a reference image and dividing the input reference image into blocks each including a plurality of pixels. The statistical model selection means (1203) receives the position of the pixel to be encoded, the reference block, and the statistical model table as input,
It is a means for selecting a statistical model from the statistical model table (1204) according to the states of the surrounding pixels of the reference block of the target pixel and outputting it to the entropy coding means (1205). The entropy decoding means (1201) is means for inputting encoded data and decoding the encoded data based on a statistical model to obtain a target block. The reference block use determining means (1205) is means for comparing the target block with the reference block and switching the subsequent processing.
The target pixel decoding means (605) is a means for decoding the encoded data to obtain a target block.

The operation of the image decoding apparatus of the present embodiment configured as described above will be described below, and the operation of one embodiment of the image decoding method of the present invention will be described at the same time.

First, the entropy decoding means (120
1) is equivalent to the entropy decoding means (801). The blocking means 2 (1202) is equivalent to the blocking means 2 (102). Statistical model selection means (12
03) is equivalent to the statistical model selection means (703). The statistical model table (1204) is equivalent to the statistical model table (704). The reference block use control means (1205) is a reference block use control means (60).
It is equivalent to 4). Target pixel decoding means (1206)
Is equivalent to the target pixel decoding means (605).

As described above, according to the present embodiment, the reference block use control means changes the coding method for blocks that do not match the statistical model, thereby reducing the blocks with poor coding efficiency and reducing the coding. A small amount of efficient decoding is possible. (Embodiment 13) FIG. 13 is a block diagram showing a configuration of an image coding apparatus according to an embodiment of the present invention, and the configuration of the present embodiment will be described with reference to FIG.

In the figure, blocking means 1 (130
1) is a means that receives an image to be encoded as an input and divides the input image into blocks each including a plurality of pixels. The blocking unit 2 (1302) is a unit that receives a reference image as an input and divides the input reference image into blocks each including a plurality of pixels. Statistical model estimation means (130
3) is means for estimating the statistical model of the target block from the reference block and storing it in the statistical model (1304). The entropy coding means (1305) is means for coding the pixels of the target block based on the statistical model (1304) and outputting the coded data.

The operation of the image coding apparatus of the present embodiment configured as described above will be described below, and the operation of one embodiment of the image coding method of the present invention will be described at the same time.

That is, the blocking means 1 (1301)
It is equivalent to the blocking means 1 (101). The blocking means 2 (1302) is equivalent to the blocking means 2 (102). The statistical model estimation means (1303) estimates a statistical model from the reference block. FIG. 29 is an explanatory diagram of the statistical model estimation method of the statistical model estimation means (1303).

In the statistical model estimation, the frequency Z of the symbol 0 is first obtained. The frequency Z is obtained by counting the number of symbols 0 in the reference block and dividing it by the total number of pixels 64. Frequency Z
From the occurrence probability of symbol 0 to the conversion graph (260
1) is used. In the conversion graph, r = 0.1. The statistical model to be estimated using the occurrence probability z obtained from the transformation graph is [0, z) as the symbols 0, [z,
1.0) is the symbol 1. The estimated statistical model is stored in the statistical model (1304).

The entropy coding means (1305)
Similar to the entropy coding means (102), the target block is coded using the statistical model (1304) estimated as the arithmetic coder.

As described above, according to the present embodiment, the statistical model estimation means estimates the statistical model of the symbols in the target block from the reference block, thereby improving the efficiency of entropy coding and reducing the code amount. Less efficient encoding is possible.

In this embodiment, the statistical model is generated for each target block, but the present invention is not limited to this. For example, a statistical model may be generated for each target pixel. (Embodiment 14) FIG. 14 is a block diagram showing a configuration of an image decoding apparatus according to an embodiment of the present invention, and the configuration of the present embodiment will be described with reference to FIG.

In the figure, blocking means 2 (140
2) is a means for inputting a reference image and dividing the input reference image into blocks each including a plurality of pixels. The statistical model estimation means (1403) estimates the statistical model of the target block from the reference block, and the statistical model (1404)
It is a means to store in. The entropy decoding means (1401) is means for receiving the encoded data as input and decoding the encoded data based on the statistical model (1404) to obtain the target block.

The operation of the image decoding apparatus of the present embodiment configured as described above will be described below, and the operation of one embodiment of the image decoding method of the present invention will be described at the same time.

That is, the blocking means 2 (1402)
It is equivalent to the blocking means 2 (102). The statistical model estimating means (1403) is a statistical model estimating means (130).
It is equivalent to 3). Entropy decoding means (140
1) is decoded by using the arithmetic model decoder and the statistical model (1404) estimated by the statistical model estimating means (1403) similarly to the exclusive OR decoding means (201),
Get the target block.

As described above, according to the present embodiment, the statistical model estimating means estimates the statistical model of the symbols in the target block from the reference block, thereby improving the efficiency of entropy coding and reducing the code amount. Less efficient decoding is possible. (Embodiment 15) FIG. 15 is a block diagram showing a configuration of an image coding apparatus according to an embodiment of the present invention, and the configuration of the present embodiment will be described with reference to FIG.

In the figure, blocking means 1 (150
1) is a means that receives an image to be encoded as an input and divides the input image into blocks each including a plurality of pixels. The motion estimating means (1506) is means for searching a block similar to the target block in the reference image and obtaining the motion vector thereof.

Motion compensation block forming means 2 (1502)
Is a means for inputting a reference image and motion information, and dividing the input reference image into blocks composed of a plurality of pixels based on the motion information. Statistical model estimation means (1503)
Is a means for estimating the statistical model of the target block from the reference block and storing it in the statistical model (1504). The entropy coding means (1505) is means for coding the pixels of the target block based on the statistical model (1504) and outputting the coded data.

The operation of the image coding apparatus of the present embodiment configured as described above will be described below, and the operation of one embodiment of the image coding method of the present invention will be described at the same time.

That is, the blocking means 1 (1501) is
It is equivalent to the blocking means 1 (101). The motion compensation blocking means 2 (1502) is equivalent to the motion compensation blocking means 2 (302). Statistical model estimation means (1
504) is equivalent to the statistical model estimating means (1303). The entropy coding means (1503) is equivalent to the entropy coding means (1305). The motion estimation means (1506) is equivalent to the motion estimation means (305).

As described above, according to this embodiment, it is possible to increase the accuracy of statistical model estimation by using the motion estimation means and the motion compensation block formation means, and to perform efficient encoding with a small code amount. (Embodiment 16) FIG. 16 is a block diagram showing a configuration of an image decoding apparatus according to an embodiment of the present invention, and the configuration of the present embodiment will be described with reference to FIG.

In the figure, the motion compensation block forming means 2
(1602) is a means for inputting a reference image and motion information, and dividing the input reference image into blocks composed of a plurality of pixels based on the motion information. The statistical model estimating means (1603) is means for estimating the statistical model of the target block from the reference block and storing it in the statistical model (1604). Entropy decoding means (1601)
Is a statistical model (1604) using the encoded data as input.
Is a means for decoding encoded data to obtain a target block.

The operation of the image decoding apparatus of the present embodiment configured as described above will be described below, and the operation of one embodiment of the image decoding method of the present invention will be described at the same time.

That is, the motion compensation block forming means 2 (160
2) is equivalent to the motion compensation blocking means 2 (402). The statistical model estimating means (1603) is equivalent to the statistical model estimating means (1303). The entropy decoding means (1601) is an entropy decoding means (14
It is equivalent to 01).

As described above, according to the present embodiment, it is possible to increase the estimation accuracy of the statistical model by using the motion compensation block formation means 2 and perform efficient decoding with a small code amount. (Embodiment 17) FIG. 17 is a block diagram showing a configuration of an image coding apparatus according to an embodiment of the present invention, and the configuration of the present embodiment will be described with reference to FIG.

In the figure, blocking means 1 (170
1) is a means that receives an image to be encoded as an input and divides the input image into blocks each including a plurality of pixels. The blocking unit 2 (1702) is a unit that receives a reference image as an input and divides the input reference image into blocks each including a plurality of pixels. Statistical model estimation means (170
3) is means for estimating the statistical model of the target block from the reference block and storing it in the statistical model (1704). The entropy decoding means (1701) receives the encoded data as an input, and based on the statistical model (1704),
It is a means for decoding the encoded data to obtain the target block. The reference block usage determining means (1706) is means for comparing the target block with the reference block, outputting a reference block usage determining signal, and switching the subsequent processing.

The operation of the image coding apparatus of the present embodiment configured as described above will be described below, and the operation of one embodiment of the image coding method of the present invention will be described at the same time.

That is, the blocking means 1 (1701) is
It is equivalent to the blocking means 1 (101). The blocking means 2 (1702) is equivalent to the blocking means 2 (102). The statistical model estimating means (1703) is equivalent to the statistical model estimating means (1303). The entropy coding means (1705) is equivalent to the entropy coding means (1305). The reference block use determination means (1706) is a reference block use determination means (505).
Is equivalent to The target pixel decoding means (1707) is equivalent to the target pixel decoding means (605).

As described above, according to the present embodiment, the reference block utilization control means changes the coding method for blocks that do not match the statistical model, thereby reducing the blocks with poor coding efficiency and A small amount of efficient decoding is possible. (Embodiment 18) FIG. 18 is a block diagram showing a configuration of an image decoding apparatus according to an embodiment of the present invention, and the configuration of the present embodiment will be described with reference to FIG.

In the figure, blocking means 2 (180
2) is a means for inputting a reference image and dividing the input reference image into blocks each including a plurality of pixels. The statistical model estimation means (1803) estimates the statistical model of the target block from the reference block, and the statistical model (1804)
It is a means to store in. The entropy decoding means (1801) is a means for receiving the encoded data and decoding the encoded data based on the statistical model (1804) to obtain the target block. The reference block usage control means (1805) is a means for switching the subsequent processing according to the reference block usage determination signal. The target pixel encoding means (1806) is means for encoding the target block and outputting the encoded data.

The operation of the image decoding apparatus of the present embodiment configured as described above will be described below, and the operation of one embodiment of the image decoding method of the present invention will be described at the same time.

That is, the blocking means 2 (1802)
It is equivalent to the blocking means 2 (102). The statistical model estimating means (1803) is a statistical model estimating means (130).
It is equivalent to 3). Entropy decoding means (180
1) is equivalent to the entropy decoding means (1401). The reference block use control means (1805) is equivalent to the reference block use control means (604). The target pixel decoding means (1806) is a target pixel decoding means (6
It is equivalent to 05).

As described above, according to the present embodiment, the reference block use control means changes the coding method for blocks that do not match the statistical model, thereby reducing the blocks with poor coding efficiency and A small amount of efficient decoding is possible.

On the other hand, a magnetic recording medium or an optical recording medium recording a program for causing a computer to execute the functions of all or some of the means described in any one of the above-described embodiments. Also, the same effect as described above can be achieved by creating the above and causing the computer to execute the above operation.

As described above, in the image coding device, the image decoding device, the image coding method, and the image decoding method of the present invention, the conventional binary image coding technique and binary image coding technique are used for the following reasons. It enables more efficient encoding and decoding than using image decoding technology. That is, (1) the target image is predicted from the reference image in the binary moving image, and the residual is expressed by exclusive OR. (2) An appropriate statistical model is always used by switching the statistical model according to the states of surrounding pixels of another frame having a correlation. (3) By creating a statistical model from the reference image,
Use an appropriate statistical model. (4) The number of blocks that do not fit the statistical model is reduced by changing the coding method using a threshold value that uses motion compensation or the sum of absolute value differences.

The reference block use determining means of the present invention has been described in the fifth embodiment with respect to the configuration shown in FIG. 5, but the present invention is not limited to this and may have the configuration shown in FIG. 30, for example. That is, in the case of the image encoding device shown in FIG.
The reference block usage determining unit 3505 is a unit that compares the coding amount of the target block with the coding amount of the reference block, outputs a reference block usage determining signal, and switches subsequent processes. This is the main difference from the configuration of FIG. As a result, the reference block use determining unit 3505 uses the encoding amount in the target image encoding unit 506 and the encoding amount in the exclusive OR encoding unit 504 to perform encoding in the target image encoding unit 506. If the amount is smaller, the encoded data is output, and if the amount of code in the exclusive OR encoding means 504 is smaller, the encoded data is output. If the code amount in the target image coding means 506 is smaller, the reference block utilization determination signal is output. As described above, according to the present embodiment, the reference block use determining unit 3505 uses the code amount as a reference and switches the coded data to reduce the blocks with poor coding efficiency and reduce the code amount. Fewer and more efficient encodings are possible. Note that the same reference numerals are given to those basically the same as the configuration of FIG.

Further, although the reference block use control means of the present invention has been described in the sixth embodiment with respect to the configuration shown in FIG. 6, the configuration is not limited to this and may be the configuration shown in FIG. 31, for example. That is, the image decoding apparatus shown in FIG.
Alternatively, based on a target pixel decoding unit 605 that obtains a target block by decoding the coded data output from the image coding apparatus illustrated in FIG. 30, and the reference block use determination signal output from the image coding apparatus, This is a configuration including a reference block use control unit 3604 that causes the output from the target block configuration unit 603 or the output from the target pixel decoding unit 605 to be output as a target block. As a result, decoding can be performed more efficiently than in the past. Note that the same reference numerals are given to those basically the same as the configuration of FIG.

In the seventh embodiment, the statistical model selecting means of the present invention has been described for the case where the statistical model selection result is not output to the decoding device. However, the present invention is not limited to this. As shown in FIG.
703 may be means for outputting the selection result of the statistical model as a selection result signal to the decoding device side. Incidentally, FIG.
Components that are basically the same as the configuration of 1 are assigned the same reference numerals.

Further, in the case where the image decoding apparatus of the present invention is a decoding apparatus corresponding to the image coding apparatus of the configuration which does not output the selection result of the statistical model to the decoding apparatus side in the above-mentioned Embodiment 8. Although described, the configuration is not limited to this and may be the configuration shown in FIG. That is, the image decoding apparatus shown in FIG. 33 obtains the selection result signal output from the image encoding apparatus shown in FIG. 32, and selects a statistical model corresponding to the selection result signal from a plurality of statistical models. It is provided with a selection unit 3803 and an entropy decoding unit 801 that entropy-decodes the encoded data output from the image encoding device based on the selected statistical model to obtain a target block. As a result, decoding can be performed more efficiently than in the past. Further, in the case of the present embodiment, the blocking means 2 (802) as shown in FIG.
Is unnecessary. In addition, the same reference numerals are given to those basically the same as the configuration of FIG.

In the eleventh embodiment, the reference block use determining means of the present invention has been described with respect to the configuration shown in FIG. 11, but the configuration is not limited to this, and the configuration shown in FIG. 34 may be used. That is, in the case of the image coding apparatus shown in FIG. 34, the reference block use determining unit 3106 compares the coding amount of the target block with the coding amount of the reference block,
It is a means for outputting a reference block use determination signal and switching the subsequent processing. This is the main difference from the configuration of FIG. As a result, the reference block use determining unit 3
In 106, if the coding amount in the target image coding unit 1107 is smaller than the coding amount in the target image coding unit 1107 and the coding amount in the entropy coding unit 1105, the coding is performed. When the data is output and the code amount in the entropy coding unit 1105 is smaller, the coded data is output. If the code amount in the target image encoding means 1107 is smaller, the reference block utilization determination signal is output. As described above, according to the present embodiment, the reference block usage determining unit 31106 uses the code amount as a reference and switches the coded data to reduce the blocks with low coding efficiency and reduce the code amount. Fewer and more efficient encodings are possible. Note that the same reference numerals are given to those basically the same as the configuration of FIG.

Further, although the reference block use control means of the present invention has been described in the above-mentioned Embodiment 12 in the case of the configuration shown in FIG. 12, the configuration is not limited to this, and the configuration shown in FIG. That is, the image decoding apparatus shown in FIG.
The target pixel decoding means 1206 for decoding the encoded data output from the image coding apparatus shown in FIG. 11 or 34 to obtain a target block, and the reference block use determination signal output from the image coding apparatus. Based on this, the reference block use control means 3205 that causes the output from the entropy decoding means 1201 or the output from the target pixel decoding means 1206 to be output as a target block.
It has a configuration including and. As a result, decoding can be performed more efficiently than in the conventional case. Note that the same reference numerals are basically attached to the same components as those in FIG.

In the above embodiment, the moving image t + 1
Use the frame binary as the target binary image and refer to the t frame
Although the case of using the value image has been described, the present invention is not limited to this. For example, the same subject is photographed by a stereo camera, and the images photographed by one camera at the same time are targeted.
The value image may be used and the image captured by the other camera may be used as the reference binary image. Even in this case, the same effect as described above is exhibited.

As described above, the first aspect related to the present invention
In this technique, a target binary image that is an image to be coded is input, the target binary image is divided into blocks including a plurality of pixels, and a block forming unit 1 that obtains the target block and a reference obtained in advance are provided. The binary image is divided into blocks including a plurality of pixels, the block forming means 2 for obtaining a reference block, the target block and the reference block are sequentially scanned, the exclusive OR of both pixel values is detected, and the exclusive OR is obtained. The image encoding device includes an exclusive-or block forming unit that forms a block, and an exclusive-or encoding unit that generates a code string of the exclusive-OR and outputs it as encoded data.

A second technique related to the present invention is to divide a pre-obtained reference binary image into blocks including a plurality of pixels to obtain a reference block, and a block forming means 2 described above. An exclusive OR decoding unit for decoding the coded data encoded by the image encoding device of the technology to obtain an exclusive OR block, and the exclusive OR block and the reference block are combined to form a target block. An image decoding apparatus including a target block configuration unit.

A third technique related to the present invention receives a target binary image which is an image to be coded, divides the target binary image into blocks including a plurality of pixels, and obtains the target block. Blocking means 1 and reference 2 obtained in advance
Blocking means 2 for dividing a value image into blocks including a plurality of pixels to obtain a reference block, and a state of pixels around the reference pixel in a reference block including a reference pixel corresponding to a target pixel in the target block An image including: a statistical model selecting unit that selects a statistical model from a plurality of statistical models; and an entropy encoding unit that entropy-encodes the target pixel based on the selected statistical model and outputs encoded data. It is an encoding device.

Further, a fourth technique related to the present invention is to divide a previously obtained reference binary image into blocks including a plurality of pixels to obtain a reference block, and a block forming means 2 and the target block. A statistical model selecting means for selecting a statistical model from a plurality of statistical models based on states of pixels surrounding the reference pixel in a reference block including a reference pixel corresponding to the target pixel; and based on the selected statistical model. An image decoding apparatus including: an entropy decoding unit that entropy-decodes encoded data output from the image encoding apparatus of the third technique to obtain a target block.

Further, in the fifth technique related to the present invention, the block most similar to the target block is referred to in the reference 2
The motion compensation block further includes a motion estimation unit that searches for a value image and obtains motion information from the search result, and the blocking unit 2 motion-compensates the reference binary image with the motion information to obtain a reference block. The image encoding device according to the third technique, which is the encoding unit 2 and outputs the motion information.

Further, in a sixth technique related to the present invention, the blocking means 2 outputs the previously obtained reference binary image from the image encoding device of the fifth technique. The image decoding apparatus according to the above-mentioned fourth technique, which is a motion-compensation block forming unit 2 that obtains a reference block by performing motion compensation based on motion information.

The seventh technique related to the present invention compares the target block with the reference block, determines whether or not to use the reference block based on the comparison result, and The reference block use determining unit refers to the reference block use determining unit that includes a reference block use determining unit that switches driving of various units, and a target pixel encoding unit that generates a coded sequence of pixel values of the target block and outputs the coded data as encoded data. When it is determined that the block is used, the entropy coding unit and the statistical model selection unit are driven to output the coded data from the entropy coding unit, and when it is determined that the block is not used, the target The pixel coding means is driven, the coded data is output from the target pixel coding means, and it is determined whether or not the reference block is used. Outputting the reference block usage determination signal is an image coding apparatus of the third technology characterized.

Further, an eighth technique related to the present invention determines whether or not to use the reference block based on a reference block use determination signal output from the image encoding device of the seventh technique. And a target pixel decoding unit that decodes the encoded data output from the image encoding device to obtain a target block, and uses the reference block. When the control means determines to use the reference block, the entropy decoding means and the statistical model selection means are driven to output the target block from the entropy decoding means, and when it is determined that the reference block is not used, The fourth step of driving the target pixel decoding means to output the target block from the target pixel decoding means
It is an image decoding device of the technology of.

The ninth technique related to the present invention is a block in which a target binary image which is an encoded image is input, and the target binary image is divided into blocks including a plurality of pixels to obtain a target block. And a blocking means 2 for dividing a previously obtained reference binary image into blocks including a plurality of pixels to obtain a reference block, and a statistical model generation for generating a statistical model of a target pixel from the reference block. And an entropy coding unit that entropy-codes the target pixel based on the generated statistical model and outputs the coded data as an entropy coding unit.

The tenth technique related to the present invention is a block forming means 2 for obtaining a reference block by dividing a reference binary image obtained in advance into blocks containing a plurality of pixels.
Statistical model generation means for generating a statistical model of the target pixel from the reference block, and coded data output from the image coding device of the ninth technique are entropy-decoded based on the generated statistical model, And an entropy decoding unit for obtaining blocks.

The eleventh technique related to the present invention uses as an input a target binary image which is a coded image, and
Dividing a value image into blocks including a plurality of pixels to obtain a target block; dividing a reference binary image obtained in advance into blocks including a plurality of pixels to obtain a reference block; And the reference block are sequentially scanned, an exclusive OR of both pixel values is detected, an exclusive OR block is formed, and a code string of the exclusive OR is generated and output as encoded data. This is a provided image encoding method.

A twelfth technique related to the present invention is to divide a reference binary image obtained in advance into blocks containing a plurality of pixels to obtain a reference block, and the first technique.
The coded data coded by the image coding method of the first technology is input, the step of decoding the coded data to obtain an exclusive OR block, and synthesizing the exclusive OR block and the reference block, And a step of configuring a target block.

The thirteenth technique related to the present invention uses as an input a target binary image which is a coded image, and
Block step 1 of dividing a value image into blocks including a plurality of pixels to obtain a target block, and block step 2 of dividing a reference binary image obtained in advance into blocks including a plurality of pixels to obtain a reference block And, based on the state of the surrounding pixels of the reference pixel in a reference block including a reference pixel corresponding to the target pixel in the target block,
Image coding including a statistical model selection step of selecting a statistical model from a plurality of statistical models, and an entropy coding step of entropy coding the target pixel based on the selected statistical model and outputting coded data. Is the way.

The fourteenth technique related to the present invention is to divide a previously obtained reference binary image into blocks including a plurality of pixels to obtain a reference block.
And a statistical model selecting step of selecting a statistical model from a plurality of statistical models based on a state of pixels surrounding the reference pixel in a reference block including a reference pixel corresponding to a target pixel in the target block; An image decoding method including an entropy decoding step of entropy-decoding encoded data output by the image encoding method of the thirteenth technique based on the statistical model to obtain a target block.

The fifteenth technique related to the present invention further comprises a motion estimation step of searching a block most similar to the target block from the reference binary image and obtaining motion information from the search result. The block formation step 2 is a motion compensation block formation step 2 in which the reference binary image is motion-compensated with the motion information to obtain a reference block, and the motion encoding is output according to the thirteenth technique. Is.

According to a sixteenth technique related to the present invention, in the blocking step 2, the previously obtained reference binary image is output by the image encoding method of the fifteenth technique. The image decoding method according to the fourteenth technique described above, which is a motion compensation block formation step 2 for obtaining a reference block by motion compensation based on motion information.

The seventeenth technique related to the present invention compares the target block with the reference block, determines whether or not to use the reference block based on the comparison result, and A reference block use determination step that switches execution of various steps, and a target pixel encoding step that generates an encoded string of pixel values of the target block and outputs the encoded string as encoded data, are referred to in the reference block use determination step. If it is determined that the block is used, the entropy coding step and the statistical model selection step are executed, the encoded data is output by the entropy coding step, and if it is determined that the block is not used, the target The pixel coding step is executed, and the coded data is output by the target pixel coding step. Is the thirteenth image coding method of the technology and outputs whether the determination result using the reference block as a reference block usage determination signal.

An eighteenth technique related to the present invention determines whether or not to use the reference block, based on a reference block use determination signal output by the image coding method of the seventeenth technique. And a target pixel decoding step of decoding the coded data output by the image coding method to obtain a target block, and using the reference block. When it is determined that the control step uses the reference block, the entropy decoding step and the statistical model selection step are executed, the target block is output by the entropy decoding step, and when it is determined that the reference block is not used, , Executing the target pixel decoding step, and executing the target pixel decoding step The outputting the block first
4 is an image decoding method of the fourth technology.

The nineteenth technique related to the present invention is a step of obtaining a target block by inputting a target binary image which is a coded image and dividing the target binary image into blocks including a plurality of pixels. Dividing a reference binary image obtained in advance into blocks including a plurality of pixels to obtain a reference block; generating a statistical model of the target pixel from the reference block; and generating the target pixel Image encoding method based on the generated statistical model and outputting as encoded data.

The twentieth technique related to the present invention is a step of dividing a reference binary image obtained in advance into blocks including a plurality of pixels to obtain a reference block, and a target pixel of the reference pixel from the reference block. Image decoding including a step of generating a statistical model, and a step of entropy-decoding coded data output by the image coding method of the nineteenth technique based on the generated statistical model to obtain a target block It is a method of conversion.

As is clear from the above description, the technique related to the present invention is a conventional technique which predicts a pixel to be encoded from a binary image having a high correlation obtained in advance and encodes the difference between them. The present invention relates to an image encoding device, an image decoding device, an image encoding method, and an image decoding method, which can perform encoding / decoding more efficiently than using the binary image encoding / decoding technology of. However, it has an advantage that encoding and decoding can be performed more efficiently than using the conventional binary image encoding technique.

[0138]

As is apparent from the above description, the technique related to the present invention has an advantage that decoding can be performed more efficiently than using the conventional binary image coding technique.

[Brief description of drawings]

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

FIG. 2 is a block diagram of an image decoding device according to a second embodiment of the present invention.

FIG. 3 is a block diagram of an image coding apparatus according to a third embodiment of the present invention.

FIG. 4 is a block diagram of an image decoding apparatus according to a fourth embodiment of the present invention.

FIG. 5 is a block diagram of an image encoding device according to a fifth embodiment of the present invention.

FIG. 6 is a block diagram of an image decoding apparatus according to a sixth embodiment of the present invention.

FIG. 7 is a block diagram of an image encoding device according to a seventh embodiment of the present invention.

FIG. 8 is a block diagram of an image decoding device according to an eighth embodiment of the present invention.

FIG. 9 is a block diagram of an image coding apparatus according to a ninth embodiment of the present invention.

FIG. 10 is a block diagram of an image decoding apparatus according to a tenth embodiment of the present invention.

FIG. 11 is a block diagram of an image coding apparatus according to an eleventh embodiment of the present invention.

FIG. 12 is a block diagram of an image decoding apparatus according to a twelfth embodiment of the present invention.

FIG. 13 is a block diagram of an image encoding device according to a thirteenth embodiment of the present invention.

FIG. 14 is a block diagram of an image decoding device according to a fourteenth embodiment of the present invention.

FIG. 15 is a block diagram of an image encoding device according to a fifteenth embodiment of the present invention.

FIG. 16 is a block diagram of an image decoding apparatus according to a 16th embodiment of the present invention.

FIG. 17 is a block diagram of an image encoding device according to a seventeenth embodiment of the present invention.

FIG. 18 is a block diagram of an image decoding device according to an eighteenth embodiment of the present invention.

FIG. 19 is a diagram of a reference image and a target image in a mask moving image.

FIG. 20 is an explanatory diagram of an exclusive OR block configuration.

FIG. 21 is an explanatory diagram of the principle of arithmetic coding.

FIG. 22 is a block diagram of arithmetic coding.

FIG. 23 is a diagram of a part of the statistical model table.

FIG. 24 is a diagram of a part of the statistical model table.

FIG. 25 is a diagram of a part of the statistical model table.

FIG. 26 is a diagram of a part of the statistical model table.

FIG. 27 is an explanatory diagram of an extrapolation reference block.

FIG. 28 is an explanatory diagram of indexes of the statistical model table.

FIG. 29 is an explanatory diagram of a frequency-occurrence probability conversion graph.

FIG. 30 is a block diagram of an image encoding device according to another embodiment of the present invention.

FIG. 31 is a block diagram of an image decoding apparatus according to another embodiment of the present invention.

FIG. 32 is a block diagram of an image encoding device according to another embodiment of the present invention.

FIG. 33 is a block diagram of an image decoding apparatus according to another embodiment of the present invention.

FIG. 34 is a block diagram of an image coding apparatus according to still another embodiment of the present invention.

FIG. 35 is a block diagram of an image decoding apparatus according to still another embodiment of the present invention.

[Explanation of symbols]

101 Blocking means 1 102 blocking means 2 103 exclusive OR block constructing means 104 exclusive OR encoding means 201 exclusive OR decoding means 202 blocking means 2 203 Target block configuration means 301 Blocking means 1 302 motion compensation blocking means 2 303 exclusive OR block construction means 304 exclusive OR encoding means 305 Motion estimation means 401 exclusive OR decoding means 402 motion compensation blocking means 2 403 Target block configuration means 501 Blocking means 1 502 Blocking means 2 503 Exclusive OR block constructing means 504 Exclusive-OR encoding means 505 Reference block usage determining means 506 Target Pixel Encoding Means 601 Exclusive-OR decoding means 602 blocking means 2 603 Target block configuration means 604 Reference block utilization control means 605 Target pixel decoding means 701 Blocking means 1 702 blocking means 2 703 Statistical model estimation means 704 Statistical model 705 Entropy coding means 801 Entropy decoding means 802 blocking means 2 803 Statistical model estimation means 804 Statistical model 901 Blocking means 1 902 Motion compensation blocking means 2 903 Statistical model estimation means 904 Statistical model 905 Entropy coding means 906 motion estimation means 1001 Entropy decoding means 1002 motion compensation blocking means 2 1003 Statistical model selection means 1004 Statistical model table 1101 Blocking means 1 1102 Blocking means 2 1103 Statistical model selection means 1104 Statistical model table 1105 Entropy coding means 1106 Reference block usage determining means 1107 Target pixel encoding means 1201 Entropy decoding means 1202 Blocking means 2 1203 Statistical model selection means 1204 Statistical model table 1205 Reference block utilization control means 1206 Target pixel decoding means 1301 Blocking means 1 1302 blocking means 2 1303 Statistical model estimation means 1304 Statistical model 1305 Entropy coding means 1401 Entropy decoding means 1402 Blocking means 2 1403 Statistical model estimation means 1404 Statistical model 1501 Blocking means 1 1502 Motion compensation block forming means 2 1503 Entropy coding means 1504 Statistical model estimation means 1505 Statistical model 1506 motion estimation means 1601 Entropy decoding means 1602 Motion compensation block forming means 2 1603 Statistical model estimation means 1604 Statistical model 1701 Blocking means 1 1702 Blocking means 2 1703 Statistical model estimation means 1704 Statistical model 1705 Entropy estimation means 1706 Reference block use determination means 1707 Target pixel encoding means 1801 Entropy decoding means 1802 Blocking means 2 1803 Statistical model estimation means 1804 Statistical model 1805 Reference block utilization control means 1806 Target pixel decoding means 1901 mask moving image 1902 Reference image 1903 Target image 1904 Reference split image 1905 Target split image 2001 Reference Block 2002 Target block 2003 Exclusive OR block 2101 number line 2102 range 2103 binary point 2104 Occurrence probability model 2105 symbol string 2201 beginning 2202 Range initialization 2203 symbol input 2204 Limited range 2205 End judgment 2206 binary point output 2207 End 2301 Statistical model table 2401 Reference block 2402 extrapolation reference block 2403 Extrapolation reference block 2501 reference block 2502 Target block 2503 Reference mask 2503 Target mask 2601 conversion graph

─────────────────────────────────────────────────── ─── Continuation of front page (58) Fields surveyed (Int.Cl. ⁷ , DB name) H04N ^7/ 24-7/68 H04N 1/41-1/419

Claims (10)

(57) [Claims] 1. Two indicating transparency information of each frame of a moving image.
In an image decoding apparatus that decodes coded data obtained by coding a value image for each block and obtains a binary image of the target block, the decoding is performed before the target frame having the target block to be decoded. Reference binary image from the converted frame
A reference block generation means for obtaining a reference block, which is an image, the reference Bro corresponding to the target pixel in the previous SL target block
Statistical pixel selection means for selecting a statistical model from a plurality of statistical models based on the state of reference pixels present in the reference pixel and the states of pixels surrounding the reference pixel, and the code based on the selected statistical model. An image decoding device, comprising: an arithmetic decoding unit for decoding the encoded data to obtain a binary image of the target block. 2. The image decoding apparatus according to claim 1, wherein the peripheral pixel of the reference pixel is a peripheral pixel located within a range of one pixel around the reference pixel. 3. The image decoding apparatus according to claim 1, wherein the peripheral pixels of the reference pixel are four pixels located on the upper, lower, left and right sides of the reference pixel. 4. The statistical model selection means, in addition to the state of the reference pixel and the states of pixels surrounding the reference pixel,
The image decoding apparatus according to claim 1, wherein the image decoding apparatus is a statistical model selection unit that selects a statistical model from a plurality of statistical models based on the states of pixels surrounding the target pixel. 5. The image decoding apparatus according to claim 4, wherein the peripheral pixels of the target pixel are located above, to the left, to the upper left, and to the upper right of the target pixel. 6. Two indicating transparency information of each frame of a moving image
An image decoding method for decoding a coded data obtained by coding a value image for each block to obtain a binary image of a target block, in which decoding is performed before a target frame having the target block to be decoded a reference binary image generating step of obtaining a reference binary image from reduction frame, the reference block generating step of obtaining a reference block containing a plurality of pixels from the reference binary image, corresponding to the target pixel in the previous SL target block To the reference block
The state of the reference pixel in the reference pixel, and the state of the surrounding pixels of the reference pixel, statistical model selection step of selecting a statistical model from a plurality of statistical models, the code based on the selected statistical model And an arithmetic decoding step of decoding the encoded data to obtain the binary image of the target block. 7. The image decoding method according to claim 6, wherein the peripheral pixel of the reference pixel is a peripheral pixel located within one pixel range around the reference pixel. 8. The image decoding method according to claim 6, wherein the peripheral pixels of the reference pixel are four pixels located on the upper, lower, left and right sides of the reference pixel. 9. The statistical model selecting step selects a statistical model from a plurality of statistical models based on the states of the surrounding pixels of the target pixel in addition to the states of the reference pixels and the surrounding pixels of the reference pixel. 9. The image decoding method according to claim 6, which is a statistical model selecting step to be selected. 10. The image decoding method according to claim 9, wherein the peripheral pixels of the target pixel are located above, to the left, to the upper left, and to the upper right of the target pixel.