KR102291585B1 - Method and Apparatus for pre-processing of HDR iamge encoding apparatus based on a single quality value, Method for encoding HDR iamge based on a single quality value - Google Patents

Abstract

Disclosed is a preprocessing method for implementing an encoding apparatus that encodes a high dynamic range (HDR) image by setting only a single image quality value. A pre-processing method according to an embodiment of the present invention includes: receiving a first image quality value to be applied to encoding a low dynamic range (LDR) image generated by applying tone mapping to the HDR image; Calculating an optimal second image quality value at which the restored image corresponding to the HDR image has an optimal image quality from among a plurality of second image quality values applicable to a residual image corresponding to a residual value between the HDR image and the LDR image ; and calculating an image quality relational expression representing a relationship between the first image quality value and the optimal second image quality value.

Description

Method and Apparatus for pre-processing of HDR iamge encoding apparatus based on a single quality value, Method for encoding HDR iamge based on a single quality value}

The present invention relates to pre-processing and encoding of an image encoding apparatus, and more particularly, to a pre-processing method and apparatus for an HDR image encoding apparatus based on a single image quality value, and an HDR image encoding method based on a single image quality value.

JPEG XT is the name of the international standard for high dynamic range (HDR) still images, which was started by ISO/IEC in 2012 for international standardization and is in progress. MPEG, which is an international standard for video encoding, is also in the process of standardizing HDR video encoding.

HDR is an abbreviation of high dynamic range, and dynamic range refers to the range of the maximum and minimum values of brightness displayed on the screen of a still image. Therefore, high dynamic range means that the range of this brightness value is wide, so it can contain rich screen information. On the other hand, LDR is an abbreviation of low dynamic range, and most images currently used in the market are low dynamic range (LDR) images with brightness values ranging from 0 to 255, that is, 8 bits of information.

In a situation where equipment capable of acquiring 12-bit and 14-bit still images has already been released and widely used in the market, JPEG XT appeared to establish an international standard for efficiently encoding and exchanging 8-bit or more still images.

The JPEG XT image encoding standard is defined to encode an HDR image in two layers: a base layer and a residual layer. In order to provide backward compatibility, an LDR image and a residual image are encoded using an existing JPEG encoding device, respectively, and stored as a single JPG file consisting of one byte stream.

In the JPEG XT image code and standard, the layer that encodes the LDR image is called the base layer, and the layer that encodes the residual image based on the residual value between the HDR image and the LDR image is called the residual layer. In order to encode these two layers, a new encoding device is not required, but encoding is performed through the ISO/IEC 10918 JPEG encoding device, which is already widely used in the market.

The JPEG encoding apparatus may designate a quality of an image to which encoding is applied, and the quality of a reconstructed image is determined according to the quality value.

As a result, the JPEG XT encoding apparatus is inconvenient to the user because the JPEG encoding apparatus is included in the base layer and the residual layer, and image quality values for encoding the base layer and the residual layer must be set respectively.

Therefore, the user can restore a restored image having the optimal quality corresponding within the single image quality only by specifying a single image quality without individually specifying the image quality for each of the base layer and the residual layer. The need to develop an encoding device that does this is emerging.

As a related prior art, US Patent Publication No. 2014-0086321 (title of the invention: HIGH DYNAMAIC RANGE CODECS, filing date: 2013.11.08) exists.

An object of the present invention is a preprocessing method and apparatus for implementing an encoding apparatus for encoding a high dynamic range (HDR) image by setting only a single image quality value, and a method for encoding a high dynamic range (HDR) image by setting only a single image quality value is to provide

A pre-processing method for implementing an encoding apparatus for encoding a high dynamic range (HDR) image by setting only a single image quality value according to an embodiment of the present invention for achieving the above object is generated by applying tone mapping to the HDR image. receiving a first image quality value to be applied to encoding a low dynamic range (LDR) image; Calculating an optimal second image quality value at which the restored image corresponding to the HDR image has an optimal image quality from among a plurality of second image quality values applicable to a residual image corresponding to a residual value between the HDR image and the LDR image ; and calculating an image quality relational expression representing a relationship between the first image quality value and the optimal second image quality value.

Preferably, the calculating of the image quality value relational expression comprises: performing DCT transformation by dividing the LDR image to be encoded into a plurality of blocks based on the first image quality value, and calculating DCT transformation coefficients for each block; performing DCT transform by dividing the residual image to be encoded into a plurality of blocks based on the optimal second quality value, and calculating DCT transform coefficients for each block; and calculating the image quality value relational expression based on the DCT transform coefficients for each block of the LDR image and the DCT transform coefficients for each block of the residual image.

Preferably, when the DCT transform coefficient for each block of the LDR image is determined based on the first image quality value, and the DCT transform coefficient for each block of the residual image is determined based on the optimal second image quality value, The image quality value relational expression may be expressed as an nth order equation (n is an integer greater than or equal to 1) between the first image quality value and the optimal second image quality value.

Preferably, the step of calculating the image quality value relational expression based on the DCT transform coefficient for each block of the LDR image and the DCT transform coefficient for each block of the residual image comprises: DCT transform coefficient for each block of the DCT-transformed LDR image calculating a first comparison value that is a comparison value between the two groups; calculating a second comparison value that is a comparison value between DCT transform coefficients for each block of the DCT-transformed residual image; and calculating the image quality value relational expression based on the first comparison value and the second comparison value.

Preferably, the first comparison value is a variance value of the DCT transform coefficients for each block of the DCT-transformed LDR image, and the second comparison value is a variance value of the DCT transform coefficients for each block of the DCT-transformed residual image. have.

Preferably, the image quality value relational expression may be defined based on Equation 2 or Equation 3.

[Equation 2]

[Equation 3]

는 상기 DCT 변환된 레지듀얼 이미지의 블럭별 DCT 변환 계수의 분산 값이고,

는 상기 DCT 변환된 LDR 이미지의 블럭별 DCT 변환 계수의 분산 값을 나타낸다. In this case, Q represents the optimal second quality value, and q represents the first image quality value,

is a variance value of DCT transform coefficients for each block of the DCT-transformed residual image,

denotes a variance value of DCT transform coefficients for each block of the DCT-transformed LDR image.

Preferably, in an embodiment of the present invention, the pre-processing method of the encoding apparatus displays the first image quality value and the optimal second image quality value corresponding to the first image quality value calculated based on the relational expression for the image quality value. The method may further include generating a table, wherein the mapping table may optionally further include a result value of quantitatively indicating a relationship between the first image quality value and the optimal second image quality value.

Preferably, the first image quality value and the second image quality value may be divided into one of a predetermined number of steps or may be displayed as quantitative values.

Preferably, the calculating of the optimal second image quality value comprises: generating an encoded LDR image by encoding the LDR image based on the first image quality value; generating a plurality of encoded residual images by encoding the dual images; generating a plurality of reconstructed images based on the encoded LDR image and the plurality of encoded residual images; and calculating a second image quality value corresponding to a restored image having the highest peak signal to noise ratio (PSNR) with the HDR image among the plurality of restored images as the optimal second image quality value.

In addition, in order to implement an encoding apparatus for encoding a high dynamic range (HDR) image by setting only a single image quality value according to an embodiment of the present invention for achieving the above object, the pre-processing unit mounted on the encoding apparatus is the HDR image an input unit for receiving a first image quality value to be applied to encoding a low dynamic range (LDR) image generated by applying tone mapping to ; Optimal for calculating an optimal second image quality value at which the restored image corresponding to the HDR image has an optimal image quality among a plurality of second image quality values applicable to a residual image corresponding to a residual value between the HDR image and the LDR image image quality calculation unit; and a relational expression calculator configured to calculate an image quality relational expression representing a relationship between the first image quality value and the optimal second image quality value.

Preferably, the relational expression calculator performs DCT transformation by dividing the LDR image to be encoded into a plurality of blocks based on the first image quality value, calculates DCT transform coefficients for each block, and the optimal second image quality value DCT transform is performed by dividing the residual image to be encoded into a plurality of blocks based on The image quality value relational expression may be calculated based on the DCT transform coefficients for each block.

Preferably, when the DCT transform coefficient for each block of the LDR image is determined based on the first image quality value, and the DCT transform coefficient for each block of the residual image is determined based on the optimal second image quality value, The image quality value relational expression may be expressed as an nth order equation (n is an integer greater than or equal to 1) between the first image quality value and the optimal second image quality value.

Preferably, the relational expression calculator calculates a first comparison value that is a comparison value between DCT transform coefficients for each block of the DCT-transformed LDR image, and a second comparison value that is a comparison value between DCT transform coefficients for each block of the DCT-transformed residual image. An operation of calculating a value may be further performed, and the image quality value relational expression may be calculated based on the first comparison value and the second comparison value.

Preferably, the first comparison value is a variance value of the DCT transform coefficients for each block of the DCT-transformed LDR image, and the second comparison value is a variance value of the DCT transform coefficients for each block of the DCT-transformed residual image. have.

Preferably, the image quality value relational expression may be defined based on Equation 1 or Equation 2.

[Equation 2]

[Equation 3]

는 상기 DCT 변환된 레지듀얼 이미지의 블럭별 DCT 변환 계수의 분산 값이고,

는 상기 DCT 변환된 LDR 이미지의 블럭별 DCT 변환 계수의 분산 값을 나타낸다. In this case, Q represents the optimal second quality value, and q represents the first image quality value,

is a variance value of DCT transform coefficients for each block of the DCT-transformed residual image,

denotes a variance value of DCT transform coefficients for each block of the DCT-transformed LDR image.

Preferably, the pre-processing apparatus according to an embodiment of the present invention generates a mapping table indicating the first image quality value and the optimal second image quality value corresponding to the first image quality value calculated based on the image quality value relational expression. The method may further include a mapping table generator to generate the mapping table, wherein the mapping table may optionally further include a result value of quantitatively indicating a relationship between the first image quality value and the optimal second image quality value.

Preferably, the first image quality value and the second image quality value may be divided into one of a predetermined number of steps or may be displayed as quantitative values.

Preferably, the optimal image quality value calculating unit comprises: a first encoder for encoding the LDR image based on the first image quality value to generate an encoded LDR image; a second encoder for encoding the residual images based on the plurality of second image quality values to generate a plurality of encoded residual images; a reconstruction unit for generating a plurality of reconstructed images based on the encoded LDR image and the plurality of encoded residual images; and a calculator configured to calculate a second image quality value corresponding to a restored image having the highest peak signal to noise ratio (PSNR) with the HDR image among the plurality of restored images as the optimal second image quality value.

In addition, the method of encoding a high dynamic range (HDR) image based on a single image quality value according to an embodiment of the present invention for achieving the above object is a low dynamic range (LDR) generated by applying tone mapping to the HDR image. ) receiving a first image quality value to be applied to image encoding; Among a plurality of second image quality values applicable to a residual image corresponding to a residual value between the HDR image and the LDR image, an optimal second image quality value at which the restored image corresponding to the HDR image has an optimal image quality and the first image quality value calculating the optimal second image quality value based on an image quality value relational expression indicating a relationship between image quality values; and encoding the LDR image and the residual image based on the first image quality value and the optimal second image quality value.

According to the present invention, if the user sets only the first image quality value to be applied to the low dynamic range (LDR) image, the optimal second image quality value to be applied to the residual image is automatically determined based on the image quality value relational expression. It has the advantage of being able to encode high dynamic range (HDR) images based only on values.

1 is a flowchart illustrating a pre-processing method of an encoding apparatus according to an embodiment of the present invention.
2 is a flowchart illustrating a method of calculating an optimal second image quality value to be applied to encoding a residual image according to an embodiment of the present invention.
3 is a flowchart illustrating a method of calculating a relational expression between a first image quality value to be applied to encoding an LDR image and an optimal second image quality value to be applied to encoding a residual image according to an embodiment of the present invention.
4 is a diagram illustrating a pre-processing apparatus mounted in an encoding apparatus according to an embodiment of the present invention.
5 is a diagram illustrating an optimal image quality value calculator according to an embodiment of the present invention.
6 is a flowchart illustrating a method of encoding an HDR image based on a single image quality value according to an embodiment of the present invention.
7 is a diagram illustrating an encoding system to which a preprocessing method of an encoding apparatus according to an embodiment of the present invention is applied.

Since the present invention can have various changes and can have various embodiments, specific embodiments are illustrated in the drawings and described in detail in the detailed description. However, this is not intended to limit the present invention to specific embodiments, and it should be understood to include all modifications, equivalents and substitutes included in the spirit and scope of the present invention. In describing each figure, like reference numerals have been used for like elements.

Terms such as first, second, A, and B may be used to describe various elements, but the elements should not be limited by the terms. The above terms are used only for the purpose of distinguishing one component from another. For example, without departing from the scope of the present invention, a first component may be referred to as a second component, and similarly, a second component may also be referred to as a first component. and/or includes a combination of a plurality of related listed items or any of a plurality of related listed items.

When a component is referred to as being “connected” or “connected” to another component, it is understood that the other component may be directly connected or connected to the other component, but other components may exist in between. it should be On the other hand, when it is said that a certain element is "directly connected" or "directly connected" to another element, it should be understood that no other element is present in the middle.

The terms used in the present application are only used to describe specific embodiments, and are not intended to limit the present invention. The singular expression includes the plural expression unless the context clearly dictates otherwise. In the present application, terms such as “comprise” or “have” are intended to designate that a feature, number, step, operation, component, part, or combination thereof described in the specification exists, but one or more other features It should be understood that this does not preclude the existence or addition of numbers, steps, operations, components, parts, or combinations thereof.

Unless defined otherwise, all terms used herein, including technical and scientific terms, have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. Terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with the meaning in the context of the related art, and should not be interpreted in an ideal or excessively formal meaning unless explicitly defined in the present application. does not

Hereinafter, preferred embodiments according to the present invention will be described in detail with reference to the accompanying drawings.

1 is a flowchart illustrating a pre-processing method of an encoding apparatus according to an embodiment of the present invention.

In operation 110, the preprocessing unit mounted in the encoding apparatus receives a first image quality value to be applied to encoding a low dynamic range (LDR) image generated by applying tone mapping to a high dynamic range (HDR) image.

In this case, the first image quality value may be divided into one of a predetermined number of steps or may be displayed as a quantitative value. For example, the first image quality value may be divided into three steps, such as upper, middle, and lower, may be a numerical value in a quantitative range such as 1 to 100, and may be input by a user.

In addition, since the tone mapping process is a process for converting an HDR image into an LDR image and is obvious to those skilled in the art, a detailed description thereof will be omitted.

Meanwhile, in the present invention, an image includes not only a still image but also an image frame of a moving image, and the present invention can be applied to both a still image and a moving image.

In step 120, the pre-processing unit optimizes the second image quality value at which the restored image corresponding to the HDR image has the optimal image quality among the plurality of second image quality values applicable to the residual image corresponding to the residual value between the HDR image and the LDR image. to calculate

In this case, the reconstructed image corresponding to the HDR image is a reconstructed image based on the encoded LDR image and the encoded residual image, and has the highest Peak Signal to Noise Ratio (PSNR) between the original HDR image and the reconstructed image. The second image quality corresponding to , becomes the optimal second image quality value.

For example, when the first image quality value corresponding to the LDR image is 'up', the second image quality value corresponding to the residual image may be automatically determined as 'up', and the restored image corresponds to the first image quality value of 'up'. It can be restored as an image.

A detailed operation of calculating the optimal second image quality value will be described later with reference to FIG. 2 .

In step 130, the pre-processing unit calculates an image quality relational expression representing a relationship between the first image quality value and the optimal second image quality value.

As described above, in the present invention, when a first image quality value to be applied to encoding an LDR image is input from a user, the image quality value relational expression for calculating an optimal second image quality value to be applied to encoding a corresponding residual image is calculated. In the case of encoding , even if only the first image quality value is input from the user, it is possible to automatically determine the optimal second image quality value based on the image quality value relational expression.

That is, according to the present invention, if the user sets only the first image quality value, the optimal second image quality value is automatically determined based on the image quality value relational expression, so that the HDR image can be encoded based on a single image quality value (the first image quality value). has the advantage of being

In addition, when HDR image encoding is performed using the image quality relational expression, the entire process of calculating the optimal second image quality value is performed even when the optimal second image quality value corresponding to the specific first image quality value is not previously calculated. It is possible to simply calculate the optimal second image quality value corresponding to the first image quality value through the image quality value relational expression without the need for a new process, thereby reducing the amount of computation and the computation time for calculating the optimal second image quality value.

For example, when the range of the first image quality value is from 1 to 100, even if only the optimal second image quality value corresponding to the first image quality value from 1 to 40 is calculated in advance, the first image quality value corresponding to the first image quality value 41 to 100 is calculated. The optimal second image quality value can be simply calculated without complicated calculation through the image quality value relational expression.

2 is a flowchart illustrating a method of calculating an optimal second image quality value to be applied to encoding a residual image according to an embodiment of the present invention.

In step 210, the preprocessor generates an encoded LDR image by encoding the LDR image based on the received first image quality value.

In operation 220, the preprocessing unit generates a plurality of encoded residual images by encoding the residual images based on the plurality of second image quality values.

For example, if the range of the second image quality value is 1 to 100, the preprocessor may generate 100 encoded residual images by encoding the residual images with the second image quality value corresponding to 1 to 100.

In step 230, the preprocessor generates a plurality of reconstructed images based on the encoded LDR image and the plurality of encoded residual images.

For example, if there are residual images encoded based on 100 second image quality values corresponding to the LDR images encoded based on the first image quality value, the preprocessor may generate 100 reconstructed images through the decoding process. have.

In operation 240, a second image quality value corresponding to a restored image having the highest peak signal to noise ratio (PSNR) with the HDR image among the plurality of restored images is calculated as an optimal second image quality value.

For example, when the first image quality value is 36, if the second image quality value corresponding to the restored image having the highest PSNR (Peak Signal to Noise Ratio) with the HDR image among 100 restored images is 52, the optimal second image quality value is 52 will become

3 is a flowchart illustrating a method of calculating a relational expression between a first image quality value to be applied to encoding an LDR image and an optimal second image quality value to be applied to encoding a residual image according to an embodiment of the present invention.

In step 310, the preprocessor divides the LDR image to be encoded into a plurality of blocks based on the first image quality value, performs DCT transform, and calculates DCT transform coefficients for each block.

More specifically, in order to encode an LDR image, DCT transformation is performed on each block of the LDR image, and encoding is performed after quantization. In step 310, DCT transformation is performed for each block for encoding the LDR image. When performed, DCT transform coefficients for each block are calculated.

In this case, the DCT transform coefficients may be calculated as different values for each block.

In operation 320, the preprocessing unit performs DCT transformation by dividing the residual image to be encoded into a plurality of blocks based on the optimal second image quality value, and calculates DCT transformation coefficients for each block.

In step 330, a picture quality value relational expression is calculated based on the DCT transform coefficients for each block of the LDR image and the DCT transform coefficients for each block of the residual image.

In this case, the image quality relational expression may be defined as an nth-order equation (n is an integer greater than or equal to 1) between the first image quality value and the optimal second image quality value, as in Equation (1).

[Equation 1]

In this case, Q denotes an optimal second image quality value, q denotes a first image quality value, and A to M are integers greater than or equal to 0.

Meanwhile, in another embodiment, the image quality relational expression may be calculated by using a comparison value between the DCT transform coefficients for each block of the LDR image and the residual image, and the following process is performed.

In a first step, the preprocessor calculates a first comparison value that is a comparison value between DCT transform coefficients for each block of the DCT-transformed LDR image.

In the second step, the preprocessor calculates a second comparison value that is a comparison value between DCT transform coefficients for each block of the DCT-transformed residual image.

In the third step, the pre-processing unit calculates an image quality relational expression based on the first comparison value and the second comparison value.

In this case, the first comparison value may be a variance value of DCT transform coefficients for each block of the DCT-transformed LDR image, and the second comparison value may be a variance value of the DCT transform coefficients for each block of the DCT-transformed residual image.

In another embodiment, the image quality value relation may be defined based on Equation 2 or Equation 3.

[Equation 2]

[Equation 3]

는 DCT 변환된 레지듀얼 이미지의 블럭별 DCT 변환 계수의 분산 값이고,

는 DCT 변환된 LDR 이미지의 블럭별 DCT 변환 계수의 분산 값을 나타낸다. In this case, Q denotes the optimal second image quality value, q denotes the first image quality value,

is the variance value of the DCT transform coefficients for each block of the DCT-transformed residual image,

denotes the variance value of the DCT transform coefficients for each block of the DCT-transformed LDR image.

Meanwhile, in another embodiment, the preprocessing apparatus may generate a mapping table indicating the first image quality value and the optimal second image quality value corresponding to the first image quality value calculated based on the relational expression of the image quality value. In this case, the mapping table may further selectively include a result value of quantitatively indicating the relationship between the first image quality value and the optimal second image quality value.

An example of a mapping table is shown in Table 1.

[Table 1]

과, DCT 변환된 LDR 이미지의 블럭별 DCT 변환 계수의 분산 값

간의 관계가 정량적으로 표시되어 있고, 하단에는 그에 대응되는 최적 제2 화질 값 Q와 제1 화질 값 q가 표시되어 있다. Referring to Table 1, the variance value of the DCT transform coefficients for each block of the DCT-transformed residual image at the top of Table 1

and the variance value of the DCT transform coefficient for each block of the DCT transformed LDR image

The relationship between the two is quantitatively displayed, and an optimal second image quality value Q and a first image quality value q corresponding thereto are indicated at the bottom.

는 0.5인 것을 알 수 있다. For example, referring to Table 1, when the first image quality value q is 36, the optimal second image quality value Q is 52

It can be seen that is 0.5.

4 is a diagram illustrating a pre-processing apparatus mounted in an encoding apparatus according to an embodiment of the present invention.

Referring to FIG. 4 , the preprocessing apparatus 400 according to an embodiment of the present invention includes an input unit 410 , an optimal image quality value calculation unit 420 , and a relational expression calculation unit 430 .

The input unit 410 receives a first quality value to be applied to encoding a low dynamic range (LDR) image generated by applying tone mapping to an HDR image.

The optimal image quality calculation unit 420 is configured to provide an optimal second image quality in which the restored image corresponding to the HDR image has the optimal image quality among a plurality of second image quality values applicable to the residual image corresponding to the residual value between the HDR image and the LDR image. Calculate the quality value.

The relational expression calculator 430 calculates an image quality relational expression representing the relationship between the first image quality value and the optimal second image quality value.

5 is a diagram illustrating an optimal image quality value calculator according to an embodiment of the present invention.

Referring to FIG. 5 , the optimum image quality value calculating unit 420 according to an embodiment of the present invention includes a first encoder 422 , a second encoder 424 , a restoration unit 426 , and an operation unit 428 . include

The first encoder 422 generates an encoded LDR image by encoding the LDR image based on the first image quality value.

The second encoder 424 generates a plurality of encoded residual images by encoding the residual images based on the plurality of second image quality values.

The reconstructor 426 generates a plurality of reconstructed images based on the encoded LDR image and the plurality of encoded residual images.

The calculator 428 calculates a second image quality value corresponding to a restored image having the highest peak signal to noise ratio (PSNR) with the HDR image among the plurality of restored images as an optimal second image quality value.

So far, a pre-processing method and apparatus of an HDR image encoding apparatus based on a single image quality value have been described. Hereinafter, a method and encoding system for encoding an HDR image based on a single image quality value will be described.

6 is a flowchart illustrating a method of encoding an HDR image based on a single image quality value according to an embodiment of the present invention.

In operation 610, the encoding apparatus receives a first quality value to be applied to encoding the LDR image.

In step 620, an optimal second image quality value for which a restored image corresponding to the HDR image has an optimal image quality among a plurality of second image quality values applicable to a residual image corresponding to a residual value between the HDR image and the LDR image, and a first image quality value An optimal second image quality value is calculated based on the image quality value relational expression indicating the relationship between the image quality values.

In operation 630, the encoding apparatus encodes the LDR image and the residual image based on the first image quality value and the optimal second image quality value.

According to the embodiment of FIG. 6 , when the encoding apparatus receives a first image quality value from a user, the LDR image and the residual image are restored so that the restored image corresponding to the HDR image has the optimal image quality based on the first image quality value. Images can be encoded.

7 is a diagram illustrating an encoding system to which a preprocessing method of an encoding apparatus according to an embodiment of the present invention is applied.

Referring to FIG. 7 , the encoding system 700 according to an embodiment of the present invention includes an HDR image storage unit 712 , a tone mapping unit 714 , an LDR image storage unit 716 , and a base layer encoding unit 718 . , a quality control unit 720 , a residual image generation unit 722 , a residual image storage unit 724 , a residual layer encoding unit 726 , and a stream output unit 728 .

The HDR image storage unit 712 stores the HDR image.

The tone mapping unit 714 generates an LDR image by applying tone mapping to the HDR image.

The LDR image storage unit 716 stores the LDR image generated by the tone mapping unit 714 .

The base layer encoder 718 receives the first image quality value q from the user and encodes the LDR image based on the first image quality value q.

The image quality control unit 720 receives the first image quality value q from the base layer encoder 718, and an optimal second image quality value Q to be applied to the residual image based on the first image quality value q. to calculate

In this case, the image quality control unit 720 may calculate the optimal second image quality value Q based on the image quality value relational expression or mapping table stored therein.

The residual image generating unit 722 generates a residual image by using a residual value between the HDR image stored in the HDR image storage unit 712 and the LDR image generated by the tone mapping unit 714 .

The residual image storage unit 724 stores the residual image generated by the residual image generation unit 720 .

The residual layer encoder 726 encodes the residual image based on the optimal second image quality value Q received from the image quality controller 720 .

The stream output unit 728 generates and outputs a stream based on the encoded LDR image and the encoded residual image generated from each of the base layer encoder 718 and the residual layer encoder 726 .

Meanwhile, the above-described embodiments of the present invention can be written as a program that can be executed on a computer, and can be implemented in a general-purpose digital computer that operates the program using a computer-readable recording medium.

The computer-readable recording medium includes a magnetic storage medium (eg, ROM, floppy disk, hard disk, etc.) and an optically readable medium (eg, CD-ROM, DVD, etc.).

So far, with respect to the present invention, the preferred embodiments have been looked at. Those of ordinary skill in the art to which the present invention pertains will understand that the present invention can be implemented in a modified form without departing from the essential characteristics of the present invention. Therefore, the disclosed embodiments are to be considered in an illustrative rather than a restrictive sense. The scope of the present invention is indicated in the claims rather than the foregoing description, and all differences within the scope equivalent thereto should be construed as being included in the present invention.

Claims (19)

A pre-processing method for implementing an encoding device that encodes a high dynamic range (HDR) image by setting only a single image quality value,
receiving a first image quality value to be applied to encoding a low dynamic range (LDR) image generated by applying tone mapping to the HDR image;
Calculating an optimal second image quality value at which the restored image corresponding to the HDR image has an optimal image quality from among a plurality of second image quality values applicable to a residual image corresponding to a residual value between the HDR image and the LDR image ; and
calculating an image quality value relational expression representing a relationship between the first image quality value and the optimal second image quality value;
Calculating the image quality value relational expression includes:
performing DCT transform by dividing the LDR image to be encoded into a plurality of blocks based on the first image quality value, and calculating DCT transform coefficients for each block;
performing DCT transformation by dividing the residual image to be encoded into a plurality of blocks based on the optimal second image quality value, and calculating DCT transform coefficients for each block; and
calculating the image quality value relational expression based on the DCT transform coefficient for each block of the LDR image and the DCT transform coefficient for each block of the residual image,
The DCT transform coefficient for each block of the LDR image is
When it is determined based on the first image quality value, and the DCT transform coefficient for each block of the residual image is determined based on the optimal second image quality value,
The image quality value relational expression is expressed as an nth-order equation (n is an integer greater than or equal to 1) between the first image quality value and the optimal second image quality value. delete delete According to claim 1,
Calculating the image quality value relational expression based on the DCT transform coefficient for each block of the LDR image and the DCT transform coefficient for each block of the residual image includes:
calculating a first comparison value that is a comparison value between DCT transform coefficients for each block of the DCT-transformed LDR image;
calculating a second comparison value that is a comparison value between DCT transform coefficients for each block of the DCT-transformed residual image; and
and calculating the image quality value relational expression based on the first comparison value and the second comparison value. 5. The method of claim 4,
The first comparison value is a variance value of DCT transform coefficients for each block of the DCT-transformed LDR image,
and the second comparison value is a variance value of DCT transform coefficients for each block of the DCT-transformed residual image. 6. The method of claim 5,
The image quality value relational expression is defined based on Equation (2) or Equation (3).
[Equation 2]

[Equation 3]

In this case, Q represents the optimal second quality value, and q represents the first image quality value,

is a variance value of DCT transform coefficients for each block of the DCT-transformed residual image,

denotes a variance value of DCT transform coefficients for each block of the DCT-transformed LDR image. 7. The method of claim 6,
The method further includes generating a mapping table indicating the first image quality value and the optimal second image quality value corresponding to the first image quality value calculated based on the image quality value relational expression;
and the mapping table further selectively includes a result value of quantitatively indicating the relationship between the first image quality value and the optimal second image quality value. According to claim 1,
The first image quality value and the second image quality value are
divided into one of a predetermined number of steps, or
A preprocessing method of an encoding device, characterized in that it is displayed as a quantitative numerical value. According to claim 1,
Calculating the optimal second image quality value includes:
generating an encoded LDR image by encoding the LDR image based on the first image quality value;
generating a plurality of encoded residual images by encoding the residual images based on the plurality of second image quality values;
generating a plurality of reconstructed images based on the encoded LDR image and the plurality of encoded residual images; and
and calculating a second quality value corresponding to a reconstructed image having the highest PSNR (Peak Signal to Noise Ratio) with the HDR image among the plurality of reconstructed images as the optimal second quality value. Method of pretreatment of the device. In the pre-processing apparatus mounted on the encoding apparatus to implement an encoding apparatus that encodes a high dynamic range (HDR) image by setting only a single image quality value,
an input unit receiving a first image quality value to be applied to encoding a low dynamic range (LDR) image generated by applying tone mapping to the HDR image;
Optimal for calculating an optimal second image quality value at which the restored image corresponding to the HDR image has an optimal image quality among a plurality of second image quality values applicable to a residual image corresponding to a residual value between the HDR image and the LDR image image quality calculation unit; and
and a relational expression calculation unit for calculating an image quality value relation representing a relationship between the first image quality value and the optimal second image quality value;
The relational expression calculation unit
DCT transformation is performed by dividing the LDR image to be encoded into a plurality of blocks based on the first image quality value, DCT transform coefficients are calculated for each block, and the residual image to be encoded based on the optimal second image quality value. is divided into a plurality of blocks to perform DCT transform, and after calculating the DCT transform coefficients for each block, based on the DCT transform coefficients for each block of the LDR image and the DCT transform coefficients for each block of the residual image, the Calculating the image quality value relation,
The DCT transform coefficient for each block of the LDR image is,
When it is determined based on the first image quality value, and the DCT transform coefficient for each block of the residual image is determined based on the optimal second image quality value,
The image quality value relational expression is expressed as an nth-order equation (n is an integer greater than or equal to 1) between the first image quality value and the optimal second image quality value. delete delete 11. The method of claim 10,
The relational expression calculation unit
Calculating a first comparison value that is a comparison value between DCT transform coefficients for each block of the DCT-transformed LDR image, and calculating a second comparison value that is a comparison value between DCT transform coefficients for each block of the DCT-transformed residual image do,
The image quality value relational expression is calculated based on the first comparison value and the second comparison value. 14. The method of claim 13,
The first comparison value is a variance value of DCT transform coefficients for each block of the DCT-transformed LDR image,
The second comparison value is a preprocessing device, characterized in that it is a variance value of DCT transform coefficients for each block of the DCT-transformed residual image. 15. The method of claim 14,
The image quality value relational expression is a pre-processing apparatus, characterized in that defined based on Equation 1 or Equation 2.
[Equation 2]

[Equation 3]

In this case, Q represents the optimal second quality value, and q represents the first image quality value,

is a variance value of DCT transform coefficients for each block of the DCT-transformed residual image,

denotes a variance value of DCT transform coefficients for each block of the DCT-transformed LDR image. 16. The method of claim 15,
and a mapping table generator for generating a mapping table displaying the first image quality value and the optimal second image quality value corresponding to the first image quality value calculated based on the image quality value relational expression;
The mapping table further comprises selectively a result value of quantitatively indicating the relationship between the first image quality value and the optimal second image quality value. 11. The method of claim 10,
The first image quality value and the second image quality value are
divided into one of a predetermined number of steps, or
A pretreatment device, characterized in that it is expressed as a quantitative number. 11. The method of claim 10,
The optimal image quality value calculation unit
a first encoder for encoding the LDR image based on the first image quality value to generate an encoded LDR image;
a second encoder for encoding the residual images based on the plurality of second image quality values to generate a plurality of encoded residual images;
a reconstruction unit for generating a plurality of reconstructed images based on the encoded LDR image and the plurality of encoded residual images; and
and a calculator for calculating a second quality value corresponding to the restored image having the highest PSNR (Peak Signal to Noise Ratio) with the HDR image among the plurality of restored images as the optimal second image quality value. Device. delete

Images