JP6069009B2 - Image decoding apparatus and image decoding program - Google Patents

Description

  The present invention relates to an image decoding apparatus and an image decoding program for decoding an image in units of blocks.

  MPEG-2, H.264. In video compression coding schemes such as H.264 / AVC (Advanced Video Coding) and HEVC (High Efficiency Video Coding) currently being standardized, each frame of the video is divided into rectangular areas called blocks for encoding. Is called. In the following, MPEG-2, H.264. H.264 / AVC and HEVC are also called image encoding techniques.

  The encoding process of the image encoding technique includes motion estimation, a process of converting an image signal into a frequency domain signal, and the like in units of blocks obtained by dividing an image (see, for example, Non-Patent Document 1).

  As a method for converting an image signal into a frequency domain signal, there are Fourier transform, wavelet transform, Hadamard transform, Karoonen-Leve transform, and the like. In image coding, DCT (discrete cosine transform) is often used.

  DCT is a process of converting a signal into a frequency domain, but can also be regarded as a process of analyzing and analyzing an original signal with a number of basis vectors (or basis images) prepared in advance.

  FIG. 1 shows an 8 × 8 basis vector. In the case of a two-dimensional signal that is an image signal, the eighth-order two-dimensional DCT represents the original two-dimensional signal with 64 types of basis vectors shown in FIG.

Supervised by Satoshi Okubo, “Revised 3rd edition H.264 / AVC textbook”, Impress R & D, p124, January 1, 2009

  In the processing of the image coding technique, the amount of information is reduced by controlling the quantization of DCT coefficients in accordance with the bit rate of transmission or recording. In general, the high frequency component of the DCT coefficient is often quantized coarsely, and the low frequency component is often finely quantized. As a result, in the DCT coefficient, there are many cases where the low frequency components remain in proportion to the high frequency components.

  For example, when image coding is performed at an extremely low bit rate, AC components other than the DC component of the DCT coefficient may become almost 0 due to quantization, and about one AC component may remain. In this case, the decoded image may be affected by one of the basis vectors shown in FIG. 1, and the pattern of the basis vector may be visually recognized as it is.

  This has a problem that it becomes a visually unpleasant image for the user who views the decoded image, leading to degradation of the image quality.

  Accordingly, an object of the present invention is to provide an image decoding apparatus and an image decoding program that can prevent deterioration in image quality by suppressing the appearance of a base vector pattern.

  An image decoding apparatus according to an aspect of the present invention is an image decoding apparatus that decodes a stream encoded using an image encoding technique, and includes frequency conversion that can be expressed by a basis vector included in the decoded stream. An appearance determination unit that determines whether or not the AC component of the frequency conversion coefficient of each encoded block can be regarded as a single coefficient for visually recognizing the base vector pattern, and the appearance determination unit visually recognizes the base vector pattern A block including a dispersion unit that distributes the maximum frequency conversion coefficient of the AC component in the block to other frequency conversion coefficients of the AC component and a distributed frequency conversion coefficient for a block determined to be regarded as a single coefficient And a reverse frequency conversion unit for performing reverse frequency conversion on the frequency conversion coefficient.

  Further, the dispersion unit may disperse the maximum frequency conversion coefficient into a frequency conversion coefficient on the DC component side.

  In addition, the distribution unit further includes a component determination unit that refers to the distribution of the frequency transform coefficients of the decoded block located above or to the left of the distribution target block, and determines whether the horizontal or vertical component is greater, The maximum frequency conversion coefficient may be dispersed in the frequency conversion coefficient on the component side determined to be large by the component determination unit.

  In addition, the appearance determination unit may be configured such that the basis vector pattern if the maximum frequency conversion coefficient is non-zero and only one, or if the ratio between the maximum frequency conversion coefficient and the second largest frequency conversion coefficient is equal to or greater than a threshold value. It may be determined that it can be regarded as a single coefficient for visual recognition.

  The frequency determination unit further determines whether the frequency of the maximum frequency conversion coefficient is higher than a predetermined frequency, and the dispersion unit has a frequency of the maximum frequency conversion coefficient higher than the predetermined frequency. If it is high, distributed processing may be performed.

  In addition, since the image decoding program according to another aspect of the present invention decodes a stream encoded using an image encoding technique, the computer performs a frequency conversion that can be expressed by a basis vector included in the decoded stream. The appearance determination step for determining whether or not the alternating current component of the frequency transform coefficient of each block encoded by the step can be regarded as a single coefficient for visually recognizing the basis vector pattern, and the appearance determination step visually recognizes the basis vector pattern. A dispersion step for distributing the maximum frequency conversion coefficient of the AC component in the block to other frequency conversion coefficients of the AC component, and a block including the distributed frequency conversion coefficient An inverse frequency conversion step of performing an inverse frequency conversion on the frequency conversion coefficient is executed.

  According to the present invention, it is possible to prevent the deterioration of image quality by suppressing the appearance of the basis vector pattern.

The figure which shows the base vector of 8x8. 1 is a block diagram illustrating an example of a schematic configuration of an image decoding device according to Embodiment 1. FIG. FIG. 3 is a block diagram illustrating an example of a configuration of a coefficient distribution unit according to the first embodiment. The figure which shows a predetermined frequency (the 1). The figure which shows a predetermined frequency (the 2). The figure which shows an example of the distribution method 1. FIG. The figure which shows an example of the distribution method 2. FIG. The figure which shows an example of the distribution method 3. FIG. 5 is a flowchart illustrating an example of image decoding processing according to the first embodiment. FIG. 6 is a block diagram illustrating an example of a configuration of an image decoding device according to a second embodiment.

  Hereinafter, each embodiment will be described in detail with reference to the accompanying drawings.

[Example 1]
The image decoding apparatus according to the first embodiment determines in advance that a base vector pattern appears in a decoded image, suppresses the appearance of a base vector pattern, and prevents deterioration in image quality.

<Configuration>
FIG. 2 is a block diagram illustrating an example of a schematic configuration of the image decoding device 10 according to the first embodiment. As illustrated in FIG. 2, the image decoding apparatus 10 includes a reception unit 101, a first decoding unit 102, an inverse quantization unit 103, a coefficient dispersion unit 104, an inverse frequency conversion unit 105, a second decoding unit 106, and a motion compensation unit 107. An adder 108 and a decoded image storage 109. An outline of each part will be described below.

  The receiving unit 101 receives a stream encoded by an image encoding technique. The receiving unit 101 outputs a stream related to the frequency transform coefficient to the first decoding unit 102 and outputs a stream related to the motion vector to the second decoding unit 106. The received stream includes header information and the like. However, in order to simplify the description, frequency conversion coefficients and motion vectors related to the present embodiment will be described below.

  The first decoding unit 102 entropy-decodes the stream acquired from the receiving unit 101 and acquires a frequency conversion coefficient. The first decoding unit 102 outputs the decoded frequency transform coefficient to the inverse quantization unit 103.

  Hereinafter, the DCT coefficient is used as an example of the frequency conversion coefficient, but the present embodiment is not limited to this. For example, in addition to DCT, DST (discrete sine transform), Hadamard transform, etc. that can be expressed by a basis vector may be used as frequency transform.

  The inverse quantization unit 103 acquires the decoded DCT coefficient, and performs inverse quantization on a block basis such as an 8 × 8 pixel size, for example. The block size is not limited to 8 × 8, but may be other sizes such as 4 × 4. The inverse quantization unit 103 outputs the inversely quantized DCT coefficient to the coefficient dispersion unit 104. For example, in the case of an 8 × 8 block, a maximum of 64 DCT coefficients are obtained.

  The coefficient distribution unit 104 determines whether or not the acquired block is a single coefficient for visually recognizing a base vector pattern when the block is decoded based on an AC component (AC component) of the obtained DCT coefficients. judge. The coefficient dispersion unit 104 disperses the DCT coefficient of the AC component of the block into another DCT coefficient of the AC component for a block that can be regarded as a single coefficient for visually recognizing the basis vector pattern. The coefficient dispersion unit 104 outputs the DCT coefficient of the block including the dispersed DCT coefficient to the inverse frequency conversion unit 105.

  The coefficient dispersion unit 104 does not perform processing on the block that is determined not to be a block that can be regarded as a single coefficient for visually recognizing the basis vector pattern, and directly uses the DCT coefficient acquired from the inverse quantization unit 103 as it is. Output to.

  The inverse frequency transform unit 105 performs inverse frequency transform (for example, inverse DCT) on the DCT coefficients of each block acquired from the coefficient dispersion unit 104 to generate an image signal or a difference image signal for motion compensation prediction. The inverse frequency conversion unit 105 outputs the generated image signal or difference image signal to the addition unit 108.

  Here, when motion compensation is performed, the second decoding unit 106 performs entropy decoding on the stream acquired from the reception unit 101 to acquire a motion vector. The second decoding unit 106 outputs the acquired motion vector to the motion compensation unit 107.

  The motion compensation unit 107 acquires a decoded reference image from a decoded image storage unit 109, which will be described later, and performs motion compensation processing using the acquired motion vector. The motion compensation unit 107 outputs the predicted image signal generated by the motion compensation process to the addition unit 108.

  When motion compensation is performed, the addition unit 108 adds the predicted image signal acquired from the motion compensation unit 107 and the difference image signal acquired from the inverse frequency transform unit 105 to generate a final decoded image. . The adder 108 generates a final decoded image based on the image signal acquired from the inverse frequency converter 105 when motion compensation is not performed. The adder 108 stores the decoded image in the decoded image storage unit 109 because it may be used in later motion compensation.

  The decoded image storage unit 109 stores the decoded image acquired from the addition unit 108 as a reference image. The reference image stored in the decoded image storage unit 109 is read out by the motion compensation unit 107.

<Coefficient dispersion unit>
Next, details of the coefficient dispersion unit in the first embodiment will be described. FIG. 3 is a block diagram illustrating an example of the configuration of the coefficient dispersion unit 104 according to the first embodiment. The coefficient distribution unit 104 illustrated in FIG. 3 includes an appearance determination unit 200, a frequency determination unit 210, a distribution unit 220, and a component determination unit 230.

  Whether the appearance determination unit 200 can regard the AC component of the frequency conversion coefficient of each block encoded by frequency conversion that can be expressed by a basis vector, included in the decoded stream, as a single coefficient that makes the basis vector pattern visible Determine whether or not.

  The appearance determining unit 200 includes a coefficient number counting unit 201, a coefficient value measuring unit 202, and a single coefficient determining unit 203 in order to realize the above processing.

  The coefficient number counting unit 201 counts the number of non-zero coefficients for the DCT coefficients of the AC component among the DCT coefficients in, for example, an 8 × 8 block. The coefficient count unit 201 outputs the count value and the non-zero DCT coefficient to the coefficient value measurement unit 202.

  The coefficient value measuring unit 202 measures the absolute value of the non-zero DCT coefficient of the acquired AC component. The coefficient value measurement unit 202 outputs the maximum DCT coefficient and the second largest DCT coefficient among the count value and non-zero DCT coefficients to the single coefficient determination unit 203.

  The single coefficient determination unit 203 uses the acquired count value, the maximum DCT coefficient, and the second largest DCT coefficient to determine whether an effective DCT coefficient other than the DC component in the block can be regarded as single (one). Determine.

  The single coefficient determination unit 203 determines that the coefficient is a single coefficient when the count value is one. The single coefficient determination unit 203 calculates the ratio of the maximum DCT coefficient and the second largest DCT coefficient when the count value is two or more. The single coefficient determination unit 203 regards the maximum DCT coefficient as a real single coefficient when this ratio is larger than the threshold (first threshold).

  The single coefficient is a DCT coefficient that can be regarded as having only one of the AC components in the block. The fact that the single coefficient is in the block means that the basis vector pattern is visually recognized in the block.

  That is, the appearance determination unit 200 determines that a block to be determined if the maximum frequency conversion coefficient is non-zero and only one, or the ratio between the maximum frequency conversion coefficient and the second largest frequency conversion coefficient is equal to or greater than the first threshold. Means that the block can be regarded as a single coefficient for visually recognizing the pattern of the base vector.

  Note that the first threshold is, for example, 10 times, and an appropriate value may be determined and set in advance by experiments or the like.

  The single coefficient determination unit 203 outputs the determined single coefficient to the frequency determination unit 210. In addition, when it is determined that there is no single coefficient, the single coefficient determination unit 203 outputs the DCT coefficient in the block to the inverse frequency conversion unit 105 as it is.

  The frequency determination unit 210 determines whether or not to perform dispersion processing based on the frequency of the single coefficient. For example, the frequency determination unit 210 determines whether the frequency of the maximum frequency conversion coefficient that is a single coefficient is higher than a predetermined frequency (second threshold).

  FIG. 4 is a diagram illustrating a predetermined frequency (part 1). In the example illustrated in FIG. 4, if a single coefficient is obtained with a basis vector other than 3 × 3 in the low frequency direction, the frequency of the single coefficient is determined to be higher than a predetermined frequency. As illustrated in FIG. 4, the frequency determination unit 210 may set a threshold value in the horizontal direction and the vertical direction as the predetermined frequency.

  FIG. 5 is a diagram showing a predetermined frequency (part 2). In the example shown in FIG. 5, if a single coefficient other than the 10 basis vectors in the low frequency direction and the upper left direction is determined, the frequency of the single coefficient is determined to be higher than a predetermined frequency. The frequency determination unit 210 may set a predetermined frequency as a boundary between a high frequency and a low frequency as shown in FIG.

  The reason why the frequency determination unit 210 performs the dispersion process only on the high-frequency single coefficient is that the low-frequency basis vector can exist in the natural image. When a high-frequency basis vector pattern that is unlikely to exist in a natural image can be visually recognized in the decoded image, it is considered that the user feels uncomfortable.

  When the single coefficient is higher than the predetermined frequency, the frequency determination unit 210 outputs the single coefficient to the dispersion unit 220. When the single coefficient is lower than the predetermined frequency, the frequency determination unit 210 performs inverse frequency conversion on the DCT coefficient in the block. Output to the unit 105. Note that the frequency determination unit 210 is not necessarily required, and the single coefficient determination unit 203 and the dispersion unit 220 may be connected.

  Returning to FIG. 3, for the block determined by the appearance determination unit 200 to be regarded as a single coefficient for visually recognizing the base vector pattern, the dispersion unit 220 calculates the maximum frequency conversion coefficient of the AC component in this block. Disperse to other frequency conversion coefficients. Further, the dispersion unit 220 may perform the dispersion process when the frequency determination unit 210 determines that the frequency of the maximum frequency conversion coefficient is higher than the predetermined frequency.

  The dispersion unit 220 includes a coefficient filter unit 221 and a normalization unit 222 in order to realize the above processing.

  The coefficient filter unit 221 performs a dispersion process (hereinafter also referred to as a filter process) only on a block including a single coefficient determined to be subjected to the dispersion process. The basic purpose of the filtering process is to disperse the single coefficient into, for example, surrounding DCT coefficients so that the basis vector cannot be visually recognized in the decoded image. Next, a method for dispersing a single coefficient will be described.

(Dispersion method 1)
The coefficient filter unit 221 disperses the single coefficient from the single coefficient to the DCT coefficient in the low frequency direction (upper left direction). FIG. 6 is a diagram illustrating an example of the distribution method 1. In the example illustrated in FIG. 6, the coefficient filter unit 221 disperses the single coefficient with respect to the DCT coefficient at the upper left of the single coefficient. The coefficient filter unit 221 disperses a predetermined percentage (for example, 50%) of the single coefficient, for example.

  As a result, the decoded image can be blurred using a low-frequency basis vector rather than a high-frequency basis vector that is highly unlikely to exist as a natural image.

  By performing the distributed processing, for example, the DCT coefficient at the upper left of the single coefficient becomes a coefficient value of a predetermined percentage of the single coefficient from 0, and the coefficient value of the single coefficient is a coefficient value obtained by subtracting a predetermined percentage from the original coefficient value. Become.

(Dispersion method 2)
The coefficient filter unit 221 may determine the dispersion direction based on the determination result of the component determination unit 230.

  The component determination unit 230 refers to the distribution of frequency transform coefficients of decoded blocks located above or to the left of the distribution target block, and determines whether there are more horizontal or vertical components. Whether there are more horizontal or vertical components in the block can be determined by the number of coefficients of the upper half and the lower half of the diagonal line of the block or the number of coefficients × the coefficient value.

  FIG. 7 is a diagram illustrating an example of the distribution method 2. In the example shown in FIG. 7, the upper block has 5 non-zero DCT coefficients in the upper half and the lower half, and the left block has 5 non-zero DCT coefficients and 2 non-zero DCT coefficients in the lower half. An example with a coefficient is shown.

  In this case, the coefficient filter unit 221 disperses the single coefficient in the upper right direction because the upper half coefficient is larger than the lower half coefficient in the upper block and the left block. At this time, the component determination unit 230 determines that there are many high-frequency components in the horizontal direction because the upper and left blocks have many upper half coefficients.

  The coefficient filter unit 221 disperses the single coefficient in the lower left direction when there are many lower half coefficients in the upper and left blocks. In addition, the coefficient filter unit 221 disperses even a little in the low frequency direction, so that when the upper half coefficient is large, the single coefficient is dispersed upward, and when the lower half coefficient is large, the single coefficient is dispersed in the left direction. Good.

  Thereby, the coefficient filter unit 221 can determine the distribution direction according to the tendency of the surrounding decoded blocks.

(Distribution method 3)
The coefficient filter unit 221 may disperse the single coefficient into a plurality of DCT coefficients instead of one DCT coefficient.

  FIG. 8 is a diagram illustrating an example of the distribution method 3. In the example illustrated in FIG. 8, the coefficient filter unit 221 disperses the single coefficient into the upper right and lower left DCT coefficients. In addition to the example shown in FIG. 8, the coefficient filter unit 221 may disperse the single coefficient into three DCT coefficients, or set the position to be dispersed to another position.

  The distribution method 3 is determined when the component determination unit 230 determines that the coefficient distribution of the upper and left blocks is uniform in the horizontal direction and the vertical direction, or when the deviation of the coefficient distribution is different between the left and upper blocks. May be applied if done.

  Returning to FIG. 3, the normalization unit 222 performs normalization processing on the DCT coefficients of the block on which the filter processing has been performed. The normalization unit 222 performs normalization so that the norm of the DCT coefficient dispersed by the filter process is equal to the norm of the DCT coefficient before dispersion. Note that the normalization unit 222 is not necessarily required as long as it is distributed in the coefficient filter unit 221 so that the norms are the same before and after the distribution.

  The above processing is performed in each block, and finally a decoded image of the entire image is output. At this time, with the above configuration, it is possible to suppress the generation of blocks that can be regarded as single coefficients for visually recognizing the pattern of the base vector, and to prevent deterioration in image quality.

<Operation>
Next, the operation of the image decoding device 10 will be described. FIG. 9 is a flowchart illustrating an example of the image decoding process according to the first embodiment. The process shown in FIG. 9 shows the process for one block.

  In step S101, the first decoding unit 102 entropy-decodes a stream related to DCT coefficients, and acquires DCT coefficients.

  In step S102, the inverse quantization unit 103 performs inverse quantization on the DCT coefficients in block units, and obtains DCT coefficients after inverse quantization.

  In step S103, the coefficient dispersion unit 104 (coefficient number counting unit 201) counts a non-zero coefficient that is an AC component among the DCT coefficients.

  In step S104, the coefficient dispersion unit 104 (single coefficient determination unit 203) determines whether the counted coefficient value is only one. If the count value is “1” (step S104—YES), the process proceeds to step S106, and if the count value is not “1” (step S104—NO), the process proceeds to step S105.

  In step S105, the coefficient dispersion unit 104 (single coefficient determination unit 203) determines that the ratio between the maximum absolute value of the DCT coefficient of the AC component and the absolute value of the second largest DCT coefficient is a threshold (first threshold). Determine if greater than. If the ratio is greater than the first threshold (step S105—YES), the process proceeds to step S106, and if the ratio is equal to or less than the first threshold (step S105—NO), the process proceeds to step S109.

  In step S106, the coefficient dispersion unit 104 (frequency determination unit 210) determines whether the frequency of the maximum DCT coefficient is higher than a threshold (second threshold). If the frequency of the maximum DCT coefficient is higher than the second threshold (step S106—YES), the process proceeds to step S107, and if the frequency of the maximum DCT coefficient is lower than the second threshold (step S106—NO), the process proceeds to step S109.

  In step S107, the coefficient dispersion unit 104 (coefficient filter unit 221) distributes the maximum DCT coefficient to nearby DCT coefficients. As the dispersion method, any one of the dispersion methods 1 to 3 may be set in the coefficient filter unit 221.

  In step S108, the coefficient dispersion unit 104 (normalization unit 222) performs normalization processing on the DCT coefficients after dispersion so that the norms of the DCT coefficients before and after dispersion are the same.

  In step S109, the inverse frequency transform unit 105 performs inverse DCT on the DCT coefficient acquired from the coefficient dispersion unit 104 to generate an image signal and a difference image signal.

  In step S110, when motion compensation is performed, the addition unit 108 adds the difference image signal and the predicted image signal, and generates a final decoded image. When the motion compensation is not performed, the adder 108 generates a final decoded image based on the image signal.

  As described above, according to the first embodiment, it is possible to prevent deterioration in image quality by suppressing the generation of blocks that can be regarded as a single coefficient for visually recognizing a base vector pattern.

[Example 2]
FIG. 10 is a block diagram illustrating an example of the configuration of the image decoding device 20 according to the second embodiment. An image decoding apparatus 20 illustrated in FIG. 10 is an example of an apparatus in which the image decoding process described in the first embodiment is implemented by software.

  As illustrated in FIG. 10, the image decoding device 20 includes a control unit 301, a main storage unit 302, an auxiliary storage unit 303, a drive device 304, a network I / F unit 306, an input unit 307, and a display unit 308. These components are connected to each other via a bus so as to be able to transmit and receive data.

  The control unit 301 is a CPU (Central Processing Unit) that performs control of each device, calculation of data, and processing in a computer. The control unit 301 is an arithmetic device that executes an image decoding program stored in the main storage unit 302 or the auxiliary storage unit 303. The control unit 301 receives data from the input unit 307 and the storage device, calculates and processes the data, and outputs the data to the display unit 308 and the storage device.

  Also, the control unit 301 can realize the processing described in the first embodiment by executing the image decoding program.

  The main storage unit 302 is a ROM (Read Only Memory), a RAM (Random Access Memory), or the like. The main storage unit 302 is a storage device that stores or temporarily stores programs and data such as OS (Operating System) and application software that are basic software executed by the control unit 301.

  The auxiliary storage unit 303 is an HDD (Hard Disk Drive) or the like, and is a storage device that stores data related to application software and the like.

  The drive device 304 reads the program from the recording medium 305, for example, a flexible disk, and installs it in the storage unit.

  In addition, an image decoding program is stored in the recording medium 305, and the program stored in the recording medium 305 is installed in the image decoding apparatus 20 via the drive device 304. The installed image decoding program can be executed by the image decoding device 20.

  The network I / F unit 306 is a peripheral having a communication function connected via a network such as a LAN (Local Area Network) or a WAN (Wide Area Network) constructed by a data transmission path such as a wired and / or wireless line. This is an interface between the device and the image decoding device 20.

  The input unit 307 includes a keyboard having cursor keys, numeric input, various function keys, and the like, a mouse and a slide pad for performing key selection on the display screen of the display unit 308, and the like. The display unit 308 is configured by an LCD (Liquid Crystal Display) or the like, and performs display according to display data input from the control unit 301.

  2 is realized by, for example, the main storage unit 302 or the auxiliary storage unit 303. Configurations other than the decoded image storage unit 109 shown in FIG. 2 are configured as, for example, the control unit 301 and the work memory. This can be realized by the main storage unit 302.

  The program executed by the image decoding device 20 has a module configuration including each unit described in the first embodiment. As actual hardware, when the control unit 301 reads a program from the auxiliary storage unit 303 and executes it, one or more of the above-described units are loaded onto the main storage unit 302, and one or more of the units are main. It is generated on the storage unit 302.

  As described above, the image decoding process described in the first embodiment may be realized as a program for causing a computer to execute the image decoding process. The processing described in the first embodiment can be realized by installing this program from a server or the like and causing the computer to execute the program.

  It is also possible to record the program on the recording medium 305 and cause the computer or portable terminal to read the recording medium 305 on which the program is recorded to realize the above-described image decoding process. The recording medium 305 is a recording medium that records information optically, electrically, or magnetically, such as a CD-ROM, a flexible disk, or a magneto-optical disk, and information is electrically stored such as a ROM or flash memory. Various types of recording media such as a semiconductor memory for recording can be used. Further, the processing described in the above-described embodiments may be implemented in one or a plurality of integrated circuits.

  Each embodiment has been described in detail above. However, the present invention is not limited to the specific embodiment, and various modifications and changes other than the above-described modification are possible within the scope described in the claims. .

DESCRIPTION OF SYMBOLS 10, 20 Image decoding apparatus 101 Reception part 102 1st decoding part 103 Inverse quantization part 104 Coefficient dispersion | distribution part 105 Inverse frequency conversion part 200 Appearance determination part 201 Coefficient number count part 202 Coefficient value measurement part 203 Single coefficient determination part 210 Frequency determination Unit 220 Dispersion unit 221 Coefficient filter unit 222 Normalization unit 230 Component determination unit 301 Control unit

Claims (6)

An image decoding apparatus for decoding a stream encoded using an image encoding technique,
It is determined whether or not the AC component of the frequency transform coefficient of each block encoded by frequency transform that can be expressed by a base vector included in the decoded stream can be regarded as a single coefficient that makes the base vector pattern visible. An appearance determination unit;
Variance for distributing the maximum frequency conversion coefficient of the AC component in the block to other frequency conversion coefficients of the AC component for the block determined by the appearance determination unit to be regarded as a single coefficient for visually recognizing the base vector pattern And
An inverse frequency transform unit for inverse frequency transforming the frequency transform coefficients in the block including the dispersed frequency transform coefficients;
An image decoding apparatus. The dispersion unit is
The image decoding apparatus according to claim 1, wherein the maximum frequency conversion coefficient is dispersed into frequency conversion coefficients on a DC component side. A component determination unit that refers to the distribution of frequency transform coefficients of the decoded block located above or to the left of the distribution target block and determines whether there are more horizontal or vertical components,
The dispersion unit is
The image decoding device according to claim 1, wherein the maximum frequency transform coefficient is distributed to frequency transform coefficients on a component side determined to be large by the component determination unit. The appearance determination unit
If the maximum frequency conversion coefficient is non-zero and only one, or if the ratio between the maximum frequency conversion coefficient and the second largest frequency conversion coefficient is greater than or equal to a threshold value, it can be regarded as a single coefficient that makes the basis vector pattern visible. The image decoding device according to any one of claims 1 to 3. A frequency determination unit that determines whether the frequency of the maximum frequency conversion coefficient is higher than a predetermined frequency;
The dispersion unit is
The image decoding apparatus according to any one of claims 1 to 4, wherein distributed processing is performed when a frequency of the maximum frequency conversion coefficient is higher than the predetermined frequency. In order to decode a stream encoded using image encoding technology,
It is determined whether or not the AC component of the frequency transform coefficient of each block encoded by frequency transform that can be expressed by a base vector included in the decoded stream can be regarded as a single coefficient that makes the base vector pattern visible. Appearance determination step,
Variance for distributing the maximum frequency conversion coefficient of the AC component in the block to other frequency conversion coefficients of the AC component for the block determined to be regarded as a single coefficient for visually recognizing the basis vector pattern in the appearance determination step Step,
An inverse frequency transform step for inverse frequency transforming the frequency transform coefficients in the block including the dispersed frequency transform coefficients;
An image decoding program for executing