WO2009130942A1 - Image encoding device and image decoding device - Google Patents

Abstract

A parameter value such as intensity of a filter process to be applied to an image or a mode whether to apply the process itself is decided by using a difference from an original image, which enables a post-process to obtain an image of a higher quality. The encoded original image data is decoded and simultaneously subjected to the low-pass filter process, the delay process, and the Laplacian filter process before being outputted to an image adjustment unit (108). The image adjustment unit (108) mixes the image data subjected to the low-pass filter process with the delayed image data by using a first parameter and generates the first parameter by using a difference between the mixed image data and the original image data. Moreover, the image data subjected to the low-pass filter process, the image data subjected to the Laplacian process, and the delayed image data are mixed by using the first and a second parameter. The second parameter is generated by using a difference between the mixed image data and the original image data.

Description

Image encoding apparatus and image decoding apparatus

The present invention relates to an image encoding device and an image decoding device, and more particularly, to a technique related to acquisition and use of auxiliary information for improving the apparent image quality.

MPEG (Moving Picture Coding Expert Group) standardized by ISO-IEC / JTCI / SC29 and JPEG (Joint Photographic Expert Group) are methods for encoding and decoding moving image data and still image data using band compression technology. Each method is known. In these methods, the screen is divided into blocks composed of a plurality of pixels, and discrete cosine (DCT) transform, quantization, and Huffman coding, which are one of orthogonal transform methods, are performed on the data in the block. As a result, a reduction in the amount of data has been realized.

Since these methods are irreversible compression methods, the image data before compression cannot normally be reproduced from the compressed data. At this time, the degree of difference between the original image data before compression and the decoded image data after compression / expansion increases in proportion to the compression rate.

By the way, the difference in the image data is an image quality deterioration factor that appears in the form of block noise that appears like a line at the block boundary, or noise such as ringing or mosquito that appears at the edge boundary of the image.
A technique for removing such coding noise is disclosed in Patent Document 1. This is to remove noise from the decoded image data. As shown in FIG. 9, each filter by the low-pass filter processing unit 11, the delay unit 13, and the Laplacian filter processing unit 12, and the adaptive filter means at the subsequent stage. An image processing apparatus having a configuration provided. At this time, the input data input to the adaptive filter means is image data to which each filter is applied.

The adaptive filter means classifies the level data stepwise by the level comparator 17 based on the level data of the high frequency component obtained by the Laplacian filter in the decoded image data. Next, the addition parameter table 19 selects coefficients (K, M) for multiplying the data to which the low-pass filter is applied, the data to which the Laplacian filter is applied, and the data delayed by the delay unit 13. Multipliers 15, 16, and 18 are multiplied, and adder 20 adds the three data.
JP-A-8-274996

The method of Patent Document 1 uses level data obtained by applying a Laplacian filter, and determines weighting parameter values (K, M) using a predetermined conversion rule. However, since the optimal value varies depending on the location of the application of this conversion rule, using a uniform value on the screen will delete necessary information or emphasize unnecessary information. Causes image quality degradation.

The present invention has been made in view of the above-described problems, and its purpose is to determine parameter values such as the strength of filter processing applied to an image and the necessity of the filter using differences from the original image. An object of the present invention is to provide an image encoding device and an image decoding device that enable post processing for obtaining a higher quality image.

In order to solve the above problems, the first technical means of the present invention encodes original image data and outputs the encoded data as encoded data, decodes the encoded data output from the encoding unit, Decoded image processing means for executing predetermined filter processing and delay processing in parallel, the decoded image is divided into a plurality of areas, and the filtered image data output from the decoded image processing means for each divided area, the delay An image quality adjustment unit that inputs processed image data and original image data, generates metadata for adjusting image quality, and outputs the metadata, and the image quality adjustment unit includes the filtered image data and A parameter generating unit that generates a parameter for mixing the delayed image data, and the parameter generating unit is an image obtained by mixing the filtered image data and the original image data using the parameters; And data based on the difference between the original image data is obtained by said generating a parameter.

According to a second technical means, in the first technical means, the decoded image processing means executes a low-pass filter process and a Laplacian filter process in parallel as the predetermined filter process, and the low-pass filter process is performed by the filter process and the delay process. The image data, the Laplacian processed image data, and the delayed image data are output, and the image quality adjustment unit mixes the low pass filter processed image data and the delayed image data for each divided region using the first parameter, and the mixed image data and the original image data are mixed. The first parameter is generated using the difference from the image data, and the low-pass filter processed image data, the Laplacian processed image data, and the delayed image data are mixed using the first and second parameters, and the mixed image data The second parameter is generated using the difference between the original image data and the original image data It is obtained by the features and.

According to a third technical means, in the second technical means, when the image quality adjustment unit generates the first parameter for each divided region, the original image and the image data obtained by mixing the low-pass filtered image data and the delayed image data Using the difference with the data, an evaluation value indicating whether the edge region indicated by the Laplacian-processed image data is present at the boundary of the block divided by the encoding unit is generated, and the first value is obtained until the evaluation value is minimized. The process of generating the parameters is repeated.

According to a fourth technical means, in the second technical means, the image quality adjustment unit mixes the low-pass filter processed image data, the Laplacian processed image data, and the delayed image data when generating the second parameter for each divided region. Using the difference between the processed image data and the original image data, an evaluation value indicating whether or not the edge region indicated by the Laplacian-processed image data exists inside the block divided by the encoding unit is generated. The process of generating the second parameter is repeated until it becomes the minimum.

According to a fifth technical means, in the third or fourth technical means, an evaluation value indicating whether or not the metadata output from the image quality adjustment unit exists at a block boundary and an evaluation indicating whether or not an edge region exists. Based on the values, the first and second parameters, the data indicating the processing contents of the low-pass filter processing and the Laplacian filter processing determined from the first parameter and the second parameter, and the region designation data for performing the filter processing are combined. It is characterized by being generated.

According to a sixth technical means, in any one of the first to fourth technical means, the metadata output from the image quality adjustment unit is data obtained by encoding a difference between decoded image data obtained by decoding encoded data and original image data. It is characterized by including.

The seventh technical means inputs the encoded data and metadata output from the image encoding apparatus as any one of the first to fifth technical means, decodes the input encoded data, and outputs the decoded image data. An image decoding apparatus comprising: a decoding processing unit to obtain; an image quality improvement processing unit that performs image quality improvement processing using decoded image data from the decoding processing unit and input metadata; A metadata decoding unit that decodes the metadata, and a filter processing unit that performs a predetermined filtering process on a specific area of the decoded image data based on the area instruction data included in the decoded metadata; A delay processing unit that delays the decoded image data, and the filtered image data and the delayed image data output from the filter processing unit and the delay processing unit, respectively, It is obtained by comprising mixing using parameters included in the data.

According to an eighth technical means, in the seventh technical means, the filter processing means executes a low pass filter process and a Laplacian filter process in parallel as the predetermined filter process, and the low pass filter process and the delay process provide a low pass filter process. Filtered image data, Laplacian processed image data, and delayed image data are output, and the low-pass filtered image data and delayed image data are multiplied by a value based on the first parameter, and the Laplacian processed image data The mixing is performed by multiplying the values based on the second parameter and adding the multiplied results.

According to the present invention, it is possible to perform post processing for obtaining a higher quality image by determining parameter values such as strength and necessity of filter processing applied to an image using differences from the original image. An image encoding device and an image decoding device can be provided. In particular, the visual image quality can be improved while taking into account the difference between the original image and the decoded image.

It is a block diagram which shows the structure of the image coding apparatus concerning one Embodiment of this invention. FIG. 2 is a block diagram illustrating a more detailed configuration of an image quality adjustment unit in FIG. 1. It is a block diagram which shows the structure of the image decoding apparatus concerning one Embodiment of this invention. FIG. 4 is a block diagram illustrating a more detailed configuration of a high image quality processing unit in FIG. 3. It is a figure which shows the example of the coefficient used for the low-pass filter process applicable to this invention, and a Laplacian filter process. It is explanatory drawing of an example of the edge emphasis process in the image coding apparatus of this invention. It is explanatory drawing of an example of the block distortion reduction process in the image coding apparatus of this invention. It is explanatory drawing of the structural example of the metadata in the image coding apparatus of this invention. It is a block diagram which shows the structural example of the conventional image processing apparatus.

Explanation of symbols

DESCRIPTION OF SYMBOLS 11 ... Low pass filter process part, 12 ... Laplacian filter process part, 13 ... Delay device, 15, 16, 18 ... Multiplier, 17 ... Level comparator, 19 ... Addition parameter table, 20 ... Adder, 101 ... Encoding part , 102 ... Decoding section, 103 ... MC section, 104 ... Delay section, 105 ... LPF section, 106 ... Delay section, 107 ... Laplacian filter section, 108 ... Image quality adjustment section, 109 ... ME section, 110 ... Image subtraction section, 111 ... Image adder, 201a, 201b, 201c ... Multiplier, 202 ... Block distortion detector, 203 ... First parameter generator, 204 ... Edge detector, 205 ... Second parameter generator, 206 ... Metadata generation 207, 208 ... Image subtracting unit, 209, 210 ... Image adding unit, 301 ... Decoding unit, 302 ... High image quality processing unit, 303 ... MC unit, 3 DESCRIPTION OF SYMBOLS 4 ... Delay part, 305 ... Image addition part, 401 ... Area selection LPF part, 402 ... Delay part, 403 ... Area selection Laplacian filter part, 404 ... Metadata decoding part, 405a-405c ... Multiplication part, 406 ... Image addition part .

Embodiments of the present invention will be described with reference to the drawings.
(Basic processing: metadata generation)
FIG. 1 is a block diagram showing a basic configuration of an image encoding apparatus according to the present invention. The apparatus includes an encoding unit 101, a decoding unit 102, an MC unit (motion compensation unit) 103, a delay unit (first delay unit) 104, an LPF unit 105, a delay unit (second delay unit) 106, and a Laplacian filter unit 107. , An image quality adjustment unit 108, an ME unit 109, an image subtraction unit 110, and an image addition unit 111.

The difference between the input original image and the decoded image one frame before is taken by the image subtracting unit 110. The encoding unit 101 divides the difference image obtained by the image subtracting unit 110 into a plurality of blocks, and performs orthogonal transformation such as DCT. Then, the coefficient obtained by the conversion is quantized, and the quantized coefficient value is converted into a variable-length code word in which a non-zero coefficient value and the number of zeros are paired.

The decoding unit 102 performs decoding processing in the reverse procedure to that of the encoding unit 101.
The image adding unit 111 adds the image compensated for motion by the MC unit 103 to the image decoded by the decoding unit 102. The image generated by the image adding unit 111 is delayed by one frame by the delay unit (first delay unit) 104.
The ME unit 109 performs motion estimation for each block from the input image and the image decoded immediately before, and generates motion vector information. The MC unit 103 performs motion compensation using the image decoded immediately before and the motion vector information generated by the ME unit 109.
The codeword generated by the encoding unit 101 and motion vector information are output as encoded data together with control information such as a mode.

The operation of the above part is, for example, MPEG-2, MPEG-4, H.264. It is also used in international standard systems such as H.264 and VC-1, and these systems can also be applied.

The decoded image generated by the image adding unit 111 is also input to the LPF unit 105, the delay unit (second delay unit) 106, and the Laplacian filter unit 107.
The LPF unit 105 filters the input decoded image using a low-pass filter. The image data generated as a result is an image obtained by removing high frequency components contained in the original image. By selectively replacing the output image data (original image) of the delay unit (second delay unit) 106 with this image, it is possible to remove the high-frequency component of that portion. As a filter coefficient used at the time of this processing, for example, the one shown in FIG. 5A can be considered, but is not limited to this.

The Laplacian filter unit 107 filters the input decoded image using a Laplacian filter. The image data generated as a result represents the distribution of high frequency components included in the image. By using this, it is possible to extract the edge of the image, and at the same time, it is possible to enhance the edge by adding the pixel value of this image and the pixel value of the output image of the second delay unit 106. . For example, the filter coefficient used in this process is shown in FIG. 5B, but is not limited thereto.

The second delay unit 106 delays the output timing of the input decoded image so as to match the output of the LPF unit 105 and the output of the Laplacian filter unit 107.
As described above, the decoded image processing means of the present invention that decodes the encoded data output from the encoding unit 101 and executes predetermined filter processing and delay processing includes the decoding unit 102, the image addition unit 111, and the LPF unit 105. The delay unit 106 and the Laplacian filter unit 107 are realized.

The image quality adjustment unit 108 combines the outputs from the LPF unit 105, the second delay unit 106, and the Laplacian filter unit 107 by a method described later, compares this with the original image, and outputs optimum combining parameters as metadata. To do.
With the configuration described above, it is possible to generate, as metadata, parameters for performing adaptive filter processing in consideration of the difference between the original image and the decoded image.

∙ Encode each block into a variable-length code. Specifically, as the encoding method, an encoding method standardized by MPEG or the like can be applied.

(Basic processing: Detailed description of the image quality adjustment unit 108)
Next, a more detailed operation of the image quality adjustment unit 108 will be described with reference to FIG.
The image quality adjustment unit 108 includes three multiplication units 201a to 201c, a block distortion detection unit 202, a first parameter generation unit 203, an edge detection unit 204, a second parameter generation unit 205, a metadata generation unit 206, and an image subtraction unit. 207 and 208, and image addition units 209 and 210.

The image quality adjustment process is performed in the following three steps.
<Step 0>: Determination of Target Area The input image data is divided into a plurality of areas (for example, M × N pixel blocks), and each area is set as an image quality adjustment process candidate area. The following <Step 1> to <Step 3> are applied to each candidate region.
<Step 1>: Determination of K The image quality adjustment unit 108 receives the LPF image from the LPF unit 105 in FIG. 1, the delay image from the delay unit 106, and the Laplacian filter image from the Laplacian filter unit 107. . An original image to be input to the image encoding device is also input.
The input LPF image and delayed image are mixed at a ratio indicated by the parameter K using the multipliers 201a and 201b and the image adder 209. This is given by the following equation where l (n) is the LPF image and d (n) is the delayed image (n is the pixel number).
(1-K) × l (n) + K × d (n)

Next, a difference image between this result and the original image is generated by the image subtracting unit 207.
The block distortion detection unit 202 generates an evaluation value indicating whether an edge region exists at a block boundary. FIG. 7 is a diagram illustrating this situation. The unit of detection is performed in units of 2 × 2 blocks, and the evaluation value is obtained by using the total value of the pixel values of the difference image at the block boundary within the image quality adjustment processing candidate area.

The first parameter generation unit 203 determines the next new parameter K based on the input evaluation value, and outputs the parameter to each of the multiplication units 201a and 201b.
These processes are repeatedly performed until K having the smallest evaluation value is found. As a method of giving K, for example, the range of K and the number of search steps are determined, and in addition to a method of squeezing, K is set to be the minimum by taking a rough initial search step in order to speed up the search time. A method such as a two-step search in which a decision is made and then a search is performed in the vicinity thereof can be considered.

<Step 2>: Determination of M The input low-pass filter image, Laplacian image, and delayed image are mixed at a ratio indicated by the parameter K and the parameter M that have already been determined using the multiplication units 201a to 201c and the image addition unit 210. Is done. This is given by the following equation, where L (n) is a Laplacian image (n is a pixel number).
(1-K) × l (n) + K × d (n) + M × L (n)

Next, a difference image between this result and the original image is generated by the image subtracting unit 208.
The edge detection unit 204 generates an evaluation value indicating whether or not the edge region indicated by the Laplacian image exists inside the block. FIG. 6 is a diagram for explaining this situation. Detection is performed in units of blocks, and an evaluation value is obtained by calculating the sum of the pixel values of the difference images inside the block inside the image quality adjustment processing candidate area.

The second parameter generation unit 205 determines the parameter M that minimizes the evaluation value, and outputs the parameter to each multiplication unit. Since the step of determining the parameter M can be performed in exactly the same manner as the step of determining the parameter K, description thereof is omitted.

<Step 3>: Generation of Metadata Through the above steps, the parameters M and K and the processing content (LPF or Laplacian) are determined for each evaluation unit. At the same time, the initial value of the evaluation value generated by the block distortion detection unit 202 and the initial value of the evaluation value generated by the edge detection unit 204 are obtained. If any of the initial values of the evaluation values is a value indicating the presence of distortion, the metadata generation unit 206 combines these to generate metadata. FIG. 8 shows a configuration example of metadata. FIG. 8A shows a format in which a set of data is simply combined. On the other hand, FIG. 8B shows an example described in XML (eXtensible Markup Language). On the other hand, the metadata about the block without any distortion is not generated.

The area instruction information in FIG. 8 represents information related to the corresponding block such as coordinate values and block numbers. The video processing name describes the processing name to be applied to the area. For example, the processing name such as “LPF” or “Laplacian” may be written as it is, or an index representing the processing name may be used. In the example described so far, the parameters are M and K values. Since what parameters are required depends on the process, the size of this area varies depending on the process name. By combining the metadata generated for each search unit, it becomes metadata for the image.

By configuring as described above, it is possible to determine parameters for performing adaptive filter processing in consideration of the difference between the original image and the decoded image.

(Basic processing: High image quality processing)
FIG. 3 is a block diagram showing the configuration of the image decoding apparatus according to the present invention.
The image decoding apparatus includes a decoding unit 301, an image quality improvement processing unit 302, an MC unit 303, a delay unit 304, and an image addition unit 305.
Next, the operation of the decoding process will be described with reference to FIG. Detailed operations of the MC unit 303 and the delay unit 304 are the same as those of the MC unit 103 and the delay unit 104 described with reference to FIG.

The decoding unit 301 decodes the variable length code included in the input encoded data, and reproduces the encoded index. Next, inverse quantization is performed to obtain transform coefficients. Finally, inverse DCT is performed to output an image. This is combined with a reference frame subjected to motion compensation, which is the output of the MC unit 303, by an image adding unit 305 to obtain a decoded image. The obtained decoded image is delayed by one frame by the delay unit 304 and then input to the MC unit 303.
Here, the decoding processing unit of the present invention that obtains decoded image data by decoding the encoded data output from the encoding device is realized by the decoding unit 301, the MC unit 303, the delay unit 304, and the image addition unit 305. ing.

The above operation is the same as that of a decoding device adopted in an international standard system such as MPEG, and therefore, MPEG-2, MPEG-4, H.264, etc. International standard systems such as H.264 and VC-1 may be used.

The high image quality processing unit 302 performs post-processing of the input decoded image using the input metadata, and outputs this as a final decoded image.

Next, a more detailed operation of the image quality improvement processing unit 302 will be described with reference to FIG. The high image quality processing unit 302 includes a region selection LPF unit 401, a delay unit 402, a region selection Laplacian filter unit 403, a metadata decoding unit 404, and multiplication units 405a to 405c. Among the above, the delay unit 402 performs the same operation as the delay unit 106 described with reference to FIG. 1, and the multiplication units 405a to 405c are the same as the multiplication units 201a to 201c described with reference to FIG. Since the operation has already been described, a detailed description of the operation is omitted.

The metadata decoding unit 404 decodes the input metadata to obtain a set of area instruction information, video processing name, and parameters.
The region selection LPF unit 401 performs the same processing as the LPF unit 105 of the image encoding device described with reference to FIG. 1 on the region specified by the region instruction information among the regions on the input decoded image. Do.
The region selection Laplacian filter unit 403 is the same as the Laplacian filter unit 107 of the image encoding device described with reference to FIG. 1 for the region specified by the region instruction information among the regions on the input decoded image. Perform processing. The region selection LPF unit 401 and the region selection Laplacian filter unit 403 correspond to the filter processing means of the present invention that executes a predetermined filter process on a specific region of the decoded image data, and the delay unit 402 converts the decoded image data into This corresponds to the delay processing means of the present invention for delaying.

The above operation is repeated for the number of sets of area instruction information, video processing name, and parameters stored in the metadata.
These processed images and the images delayed by the delay unit 402 are multiplied in units of pixels by the multiplying units 405a to 405c using the parameters K and M, respectively, and an addition process is performed in units of pixels by the image adding unit 406.

By configuring as described above, it is possible to perform image quality improvement processing with reduced edge blur and block distortion, and to limit filter processing such as LPF and Laplacian not to the entire screen but to a necessary area. Therefore, the amount of processing can be reduced, and thus the power consumption of the CPU can be reduced.

(Area specification processing)
The area instruction information has been described as information for designating a block, but an area having an arbitrary shape may be used. In this case, shape coding information is used as information for designating a region. As a shape encoding method, binary image encoding methods such as MH, MR, and MMR used in facsimiles and the like can be considered. A method using area coding standardized by MPEG-4 is also conceivable. Furthermore, a method of encoding the boundary of the region as a set of a plurality of vectors is also conceivable.
By configuring as described above, it is possible to more flexibly specify a region where image quality is improved, and it is possible to effectively improve image quality.

(Use of differential data)
In the description so far, an example of performing image quality improvement by combining a method of removing block distortion using a low-pass filter and a method of enhancing an edge has been shown, but by increasing the compression rate, distortion of the decoded image is large, If the image quality cannot be corrected by these methods, the difference between the decoded image and the original image may be encoded and used as metadata.

In this case, as a coding method, a method using a hierarchical coding technique can be considered. For example, MPEG-2 で あ れ ば SNR Scalability, MPEG-4 SNR Scalability, SVC (Scalable Video Coding) can be used for the above-mentioned international standards. In this case, the metadata may store the decoding standard name as the video processing name and the difference information encoded as the parameter.

By configuring as described above, image quality improvement that cannot be realized by image processing such as smoothing and edge enhancement can be realized, and it is possible to efficiently improve image quality by appropriately using this for conspicuous parts. Become.

Claims (8)

An encoding unit that encodes original image data and outputs the encoded image data;
Decoded image processing means for decoding the encoded data output from the encoding unit and executing predetermined filter processing and delay processing in parallel;
The decoded image is divided into a plurality of regions, and the filtered image data, the delayed image data, and the original image data output from the decoded image processing unit are input for each divided region, and the image quality is adjusted. An image quality adjustment unit that generates and outputs the metadata of
The image quality adjustment unit includes a parameter generation unit that generates a parameter for mixing the filtered image data and the delayed image data;
The parameter generation unit generates the parameter based on a difference between the original image data and image data obtained by mixing the filtered image data and the original image data using the parameters. Encoding device. The image encoding device according to claim 1,
The decoded image processing means executes a low-pass filter process and a Laplacian filter process in parallel as the predetermined filter process, and the filter process and the delay process result in low-pass filter processed image data, Laplacian-processed image data, and Output delayed image data,
The image quality adjustment unit mixes the low-pass filter processed image data and the delayed image data for each of the divided regions using a first parameter, and uses the difference between the mixed image data and the original image data. The first parameter is generated, and the low-pass filter processed image data, the Laplacian processed image data, and the delayed image data are mixed using the first and second parameters, and the mixed image data and the original image are mixed. An image encoding apparatus that generates the second parameter using a difference with data. The image encoding device according to claim 2,
The image quality adjustment unit uses the difference between the original image data and the image data obtained by mixing the low-pass filter processed image data and the delayed image data when generating the first parameter for each divided region, An evaluation value indicating whether an edge region indicated by the Laplacian-processed image data is present at a boundary between blocks divided by the encoding unit is generated, and the first parameter is generated until the evaluation value is minimized. An image encoding apparatus characterized by repeating the processing. The image encoding device according to claim 2,
When the image quality adjusting unit generates the second parameter for each of the divided areas, the original image data and the image data obtained by mixing the low-pass filter processed image data, the Laplacian processed image data, and the delayed image data Is used to generate an evaluation value indicating whether the edge region indicated by the Laplacian-processed image data exists inside the block divided by the encoding unit, and until the evaluation value is minimized An image coding apparatus characterized by repeating the process of generating the second parameter. 5. The image encoding device according to claim 3, wherein the metadata output from the image quality adjustment unit indicates an evaluation value indicating whether or not the boundary exists between the blocks and whether or not the edge region exists. Based on the evaluation value, the first and second parameters, data indicating processing contents of low-pass filter processing and Laplacian filter processing determined from the first parameter and second parameter, and region designation data for performing the filtering processing An image encoding device generated by combining the two. 5. The image encoding device according to claim 1, wherein the metadata output from the image quality adjustment unit encodes a difference between decoded image data obtained by decoding the encoded data and the original image data. An image encoding device characterized by including the processed data. A decoding processing unit that receives the encoded data and metadata output from the image encoding device according to any one of claims 1 to 5 and decodes the input encoded data to obtain decoded image data; An image quality improvement processing unit that performs image quality improvement processing using the decoded image data from the decoding processing unit and the input metadata,
The high image quality processing unit includes a metadata decoding unit that decodes the metadata;
Filter processing means for executing a predetermined filter process on a specific area of the decoded image data based on the area instruction data included in the decoded metadata;
Delay processing means for delaying the decoded image data;
An image decoding apparatus, wherein the filtered image data and the delayed image data respectively output from the filter processing means and the delay processing means are mixed using the parameters included in the decoded metadata. The image decoding device according to claim 7,
The filter processing means executes the low pass filter process and the Laplacian filter process in parallel as the predetermined filter process, and the low pass filter process image data, the Laplacian process image data, And outputting the delayed image data,
The low pass filter processed image data and the delayed image data are multiplied by a value based on a first parameter, and the Laplacian processed image data is multiplied by a value based on a second parameter, and each of the multiplications An image decoding apparatus that performs the mixing by adding the results.

Images