KR20100135638A - Method for multiple interpolation filters, and apparatus for encoding by using the same - Google Patents

Abstract

PURPOSE: A multiple interpolation filtering method and a coding device applying the same are provided to interpolate a current image by considering the characteristic of vertical or horizontal direction having circular image and to improve coding efficiency. CONSTITUTION: A multiple interpolation images filters a reference image to a plurality of interpolation filtering and to be obtained(S710). As one among the multiple interpolation filtering uses the interpolation images, if is determined as a finial interpolation filtering method(S720). When the current image is predicted based on the interpolation images, the interpolation filtering applied to interpolation image, which has high coding rate, is determined as the final interpolation filtering method. The interpolation filtering includes a separation adaptive interpolation filtering, a direction adaptive interpolation filtering, and an interpolation filtering using a filter coefficient.

Description

A plurality of interpolation filtering methods and an encoding apparatus using the same {METHOD FOR MULTIPLE INTERPOLATION FILTERS, AND APPARATUS FOR ENCODING BY USING THE SAME}

The present invention relates to an interpolation filtering method, and more particularly, to an interpolation filtering method and a filter for selectively applying interpolation filtering to a reference image.

The present invention is derived from a study conducted as part of the IT source technology development project of the Ministry of Information and Communication and the Ministry of Information and Communication Research and Development (Task Management Number: 2008-F-011-01, Task Name: Development of Next-Generation DTV Core Technology).

In the video codec, interpolation filtering is performed on a reference frame in 1/2 and 1/4 pixel units for accurate motion estimation. In particular, a method of interpolating and filtering an image using a fixed filter coefficient is mainly used because of its simplicity and excellent performance.

Recently, researches on an adaptive interpolation filter that performs interpolation filtering using adaptive filter coefficients for each image without using fixed filter coefficients are being conducted. Typically, non-separable adaptive interpolation is performed. filter, separate adaptive interpolation filter, and directional adaptive interpolation filter.

The present invention provides an interpolation filtering method and filter capable of interpolating a reference image using a plurality of interpolation filters.

In the interpolation filtering method according to an embodiment of the present invention, a plurality of interpolation images are obtained by filtering a reference image by a plurality of interpolation filtering, and using the obtained plurality of interpolation images. Determining one as the final interpolation filtering method.

In this case, in the determining, the interpolation filtering applied to the interpolation image when the encoding rate is high when the current image is respectively predicted based on the obtained plurality of interpolation images may be determined as the final interpolation filtering method.

In the determining, the interpolation filtering having the lowest computational complexity for performing interpolation filtering among the plurality of interpolation filtering may be determined as the final interpolation filtering method.

The plurality of interpolation images may include first and second interpolation images. In this case, in the obtaining, after the first interpolation image is obtained by performing vertical adaptive interpolation filtering on the reference image, the second interpolation image may be obtained by performing horizontal adaptive interpolation filtering on the reference image.

In addition, interpolation filtering information indicating the determined final interpolation filtering method may be performed.

In the interpolation filtering method according to an embodiment of the present invention, a vertical adaptive interpolation filtering of a reference image and a vertical adaptive interpolation filtering of the reference image after vertical adaptive interpolation filtering to obtain an interpolated image Steps.

According to another embodiment of the present invention, a decoding method includes determining an interpolation filtering method used in encoding based on interpolation filtering information, obtaining an interpolation image based on the determined interpolation filtering method, and based on the interpolation image. And decoding the encoded image to obtain a reconstructed image.

In this case, the method may further include receiving interpolation filtering information, and the determining may determine the interpolation filtering method based on the received interpolation filtering information. Here, the received interpolation filtering information may include at least one of an interpolation filtering indicator indicating a type of interpolation filtering applied to the final interpolation filtering method used during encoding, and a filter coefficient.

The method may further include inferring the archive filtering information from the decoded images. The determining may include determining an interpolation filtering method based on the inferred interpolation filtering information. Here, the inferred interpolation filtering information may include at least one of context information, statistical characteristics of one or more residual signals of the reconstructed image and the current image, and distribution of pixel values.

In this case, the statistical characteristic of the residual signal may include at least one of a discrete cosine value and a coded block pattern (CBP) value of the residual signal.

According to the present invention, as the reference image is interpolated by the interpolation method having the maximum coding rate, the current image may be encoded in consideration of the vertical or horizontal characteristics of the original image, and the encoding efficiency may be improved.

In addition, a reconstructed image having better image quality may be obtained by decoding an image encoded by an interpolation method having a maximum coding rate.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings. However, in describing the present invention, when it is determined that the detailed description of the related air technology or configuration may unnecessarily obscure the subject matter of the present invention, the detailed description thereof will be abbreviated or omitted.

1 is a block diagram showing a configuration of an encoding apparatus according to an embodiment of the present invention.

Referring to FIG. 1, the encoding apparatus 100 according to an embodiment of the present invention may include a motion estimator 110, a motion compensator 120, a first difference unit 125, a DCT and a quantizer 130, An entropy encoder 140, an inverse quantization and inverse DCT unit 150, a second difference unit 155, a deblocking unit 160, and a loop filter 170 are included.

The motion estimation unit 110 generates a motion vector of a preset current block in a current image to be encoded using a reference image stored in a buffer (not shown) or a storage unit (not shown). The reference picture may be a previous picture continuously input before the current picture, or a next picture continuously input after the current picture.

The motion estimator 110 may estimate the motion using various motion estimation algorithms such as block matching algorithm, phase correlation, and HSBMA.

The motion compensator 120 generates a predicted block that is a predicted value of the current block based on the motion vector generated by the motion estimator 110. In this case, the motion compensator 120 generates a prediction block by intra prediction or inter prediction.

The first difference unit 125 generates a difference block by subtracting the prediction block from the current block.

The DCT and quantization unit 130 performs discrete cosine transform on the difference block generated by the first difference unit 125 and quantizes the quantized transform coefficient.

The entropy encoder 140 entropy encodes encoding information such as a quantized transform coefficient and a motion vector and interpolation filtering information received from the interpolation filter 170 to be described later. In this case, the encoded transform coefficient and interpolation filtering information are inserted into the bitstream together with the motion vector and transmitted to the decoding apparatus 400.

The inverse quantization and inverse DCT unit 150 inverse quantizes the differential blocks quantized by the DCT and the quantization 130 in order to predict the next encoded image, and inverse discrete cosine transform to restore the differential blocks before encoding. .

The second difference unit 155 reconstructs the current block before encoding by adding the reconstructed differential block and the prediction block generated by the motion compensator 120.

The deblocking unit 160 deblocks and filters the current block reconstructed by the second differential unit 150 to output the deblocked image. As a result, the boundary error between blocks in the deblocked image is reduced to smooth the block boundary of the deblocked image.

The interpolation filter 170 interpolates and filters the reference image stored in the buffer (not shown) or the storage unit (not shown). Here, the interpolation filtering method may include at least one of separated adaptive interpolation filtering, directional adaptive interpolation filtering, adaptive interpolation filtering, and interpolation filtering using fixed coefficients. At least one of adaptive interpolation filtering and vertical reference split adaptive interpolation filtering.

A method of interpolating and filtering the deblocked image by the interpolation filter 170 will be described with reference to FIG. 2. Hereinafter, for convenience of description, among the n preset interpolation filtering methods, the first interpolation filtering will be described as the horizontal reference separation adaptive interpolation filtering, and the second interpolation filtering is the vertical reference separation adaptive interpolation filtering as an example. . Here, n represents a positive number and is preset. In this case, the first and second interpolation filtering may be different types of interpolation filtering.

 2, the interpolation filter 170 includes an interpolation unit 171 and a determination unit 173, and the interpolation unit 171 includes first interpolation units 171-1 to n-th interpolation unit 171-. n). In the present embodiment, for convenience of description, the first and second interpolators 171-1 and 171-2 will be described.

The first interpolator 171-1 obtains a first interpolated image by interpolating the reference image using a filter coefficient obtained by horizontally-referenced adaptive adaptive interpolation filtering. Here, horizontal reference separation adaptive interpolation filtering indicates that the reference image is filtered in the horizontal direction and the horizontally filtered image is filtered in the vertical direction.

Specifically, referring to FIG. 3, the first interpolation unit 171-1 may generate a, b, and the like based on pixel values of horizontal integer pixels C1 to C6 in the original image to be encoded and the corresponding reference image. c The filter coefficients of the six-tap horizontal adaptive interpolation filter corresponding to the precision pixels and the interpolation pixel values for each of the fine pixels a, b, and c are calculated.

Referring to FIG. 4, the first interpolator 171-1 may determine the precision pixels d, h, based on pixel values of the integer pixels A3 to F3 in the vertical direction in the original image to be encoded and the corresponding reference image. Calculate the filter coefficients of the six-tap vertical adaptive interpolation filter corresponding to each of them, and the interpolation pixel values for each of the fine pixels d, h, and l. Here, a and c represent 1/4 pixels and b represent 1/2 pixels, the interpolation pixel value is used as a prediction image, and the first interpolator 171-1 is a prediction error that is an error between the original image and the prediction image. The filter coefficients and interpolation pixel values are calculated to minimize.

In the same manner, referring to FIG. 4, the first interpolation unit 171-1 is applied to each of the precision pixels based on the pixel values of the vertical pixels Aa to Fa, Ab to Fb, and Ac to Fc in the vertical direction. Calculate the filter coefficients of the corresponding six-tap vertically adaptive interpolation filter and the interpolation pixel values for each of the vertical precision pixels e, i, m, the precision pixels f, j, n, and the precision pixels g, k, o. . Here, a method of calculating a filter coefficient that minimizes a prediction error and a method of interpolating a reference image using the filter coefficient are well known techniques, and thus detailed descriptions thereof will be omitted.

The second interpolator 171-2 obtains a second interpolated image by interpolating the reference image using a filter coefficient obtained by vertically-referenced adaptive adaptive interpolation filtering. Here, the vertical reference separation adaptive interpolation filtering indicates that the reference image is filtered in the vertical direction and the image filtered in the vertical direction is filtered in the horizontal direction.

Specifically, referring to FIG. 5, the second interpolation unit 171-2 may perform precision pixels d, h, based on pixel values of vertical integer pixels A3 to F3 in the original image to be encoded and the corresponding reference image. Calculate the filter coefficients of the six-tap vertical adaptive interpolation filter corresponding to each of them, and the interpolation pixel values for each of the precise pixels d, h, and l.

6, the six-tap horizontal direction corresponding to each of the precision pixels a, b, and c is based on pixel values of integer pixels C1 to C6 in the horizontal direction in the original image to be encoded and the corresponding reference image. The interpolation pixel values for each of the filter coefficients of the adaptive interpolation filter and the precision pixels a, b, and c are calculated.

In the same manner, the second interpolation unit 171-2 is formed of the horizontal precision pixels C _{1 , d} to C _{6 , d} , C _{1, h} to C _{6 , h} , and C _{1 , l} to C _{6 , l} Filter coefficients of the six-tap horizontal adaptive interpolation filter corresponding to each precision pixel based on each pixel value, and precision pixels e, f, g in the horizontal direction, precision pixels i, j, k, and precision pixels m, An interpolation pixel value for each of n and o is calculated.

The determiner 173 predicts the current image based on the first interpolated image acquired by the first interpolator 171-1 and based on the second interpolated image acquired by the second interpolator 171-2. In order to predict the current image, interpolation filtering applied to an interpolation image having a high coding rate is determined as a final interpolation filtering method.

That is, when the prediction rate of the current image based on the first interpolation image is higher than the prediction rate of the current image based on the second interpolation image, the determination unit 173 determines that the first interpolation unit 171-1 is used. In the above, the horizontal reference adaptive interpolation filtering performed to acquire the first interpolation image is determined as the final interpolation filtering method.

On the contrary, if the prediction rate of the current image based on the second interpolation image is higher than that of the prediction of the current image based on the first interpolation image, the determination unit 173 determines that the second interpolation unit 171-2 is used. In the above, the vertical reference separation adaptive interpolation filtering performed for the second interpolation image is determined as the final interpolation filtering method. Here, the method of predicting the current image using the interpolated image is various, and since it is a well known technique, a detailed description thereof will be omitted.

In addition, the determination unit 173 may calculate an error between each of the first and second interpolation images and the original image, and determine the interpolation filtering used in the interpolation image having the minimum calculated error as the final interpolation filtering method.

That is, when the error between the first interpolated image and the original image is smaller than the error between the second interpolated image and the original image, the determination unit 173 determines The horizontal interpolation-based adaptive adaptive interpolation filtering performed by the first interpolator 171-1 to acquire the first interpolation image is determined as the final interpolation filtering method. In this case, in the opposite case, if the error between the second interpolation image and the original image is smaller than the error between the first interpolation image and the original image, the determination unit 173 performs a final interpolation filtering method based on vertical reference separation adaptive interpolation filtering. Decide on

In addition, the determiner 173 encodes and decodes an error between the first interpolated image and the original image and an error between the second interpolated image and the original image, respectively, to perform final interpolation filtering on the interpolation filtering used in the interpolation image having the lowest image quality degradation. You can decide by the method.

In addition, the determination unit 173 calculates the complexity of the horizontal reference separation adaptive interpolation filtering performed by the first interpolator 171-1 to obtain the first interpolation image, and the second interpolator 171-2. The computational complexity of the vertically-referenced adaptive adaptive interpolation filtering performed to acquire the second interpolation image may be respectively calculated, and the archive filtering having a low computational complexity may be determined as the final interpolation filtering method.

The determination unit 173 generates interpolation filtering information based on the determined final interpolation filtering method, and transmits the generated interpolation filtering information to the entropy encoder 140. Then, the entropy encoder 140 entropy-codes the interpolation filtering information, inserts it into the bitstream, and transmits the entropy encoding information to the decoding apparatus 400.

Here, the interpolation filtering information includes at least one of an interpolation filtering indicator indicating a type of interpolation filtering applied to the final interpolation filtering method and a filter coefficient of interpolation filtering applied to the final interpolation filtering method. In this case, the determination unit 173 may store the generated interpolation filtering information in a storage unit (not shown).

7 is a flowchart provided to explain an interpolation filtering method according to an embodiment of the present invention.

First, the first interpolator 171-1 obtains a first interpolated image by interpolating a reference image using a filter coefficient obtained by horizontally reference-separated adaptive interpolation filtering, and obtains a second interpolator ( 171-2 obtains a second interpolated image by interpolating the reference image using a filter coefficient obtained by vertically-referenced adaptive adaptive interpolation filtering of the reference image (S710).

Specifically, as shown in FIGS. 3 and 4, when the original image has a higher frequency characteristic in the horizontal direction than in the vertical direction, the first interpolation unit 171-1 has a frequency characteristic (precision in the horizontal direction). The reference image is interpolated to more reflect the pixels a, b, and c.

5 and 6, when the original image has a higher frequency characteristic in the vertical direction than in the horizontal direction, the second interpolation unit 171-2 includes a frequency characteristic with respect to the vertical direction (precision pixel). The reference image is interpolated to more reflect d, h, l).

Here, the horizontal reference separation adaptive interpolation filtering indicates that the reference image is filtered horizontally and the horizontally filtered image is filtered vertically, and the vertical reference separation adaptive interpolation filtering indicates that the reference image is vertical. Direction filtering, and the vertically filtered image is filtered in the horizontal direction.

Next, the determination unit 173 determines a final interpolation filtering method based on the obtained first and second interpolation images (S720).

In detail, the determination unit 173 determines whether the current image is based on the first interpolated image acquired in operation S710 and when the current image is predicted based on the second interpolated image. The interpolation filtering applied to is determined as the final interpolation filtering method. Here, since a method of predicting a current image using an interpolated image is already well known, a detailed description thereof will be omitted.

In addition, the determination unit 173 may calculate an error between each of the first and second interpolation images and the original image, and determine the interpolation filtering used in the interpolation image having the minimum calculated error as the final interpolation filtering method. In this case, the determiner 173 may determine the interpolation filtering used in the interpolation image having the lowest image quality deterioration as the final interpolation filtering method by encoding and decoding the calculated error, respectively.

In addition, the determination unit 173 calculates the complexity of the horizontal reference separation adaptive interpolation filtering performed to obtain the first interpolation image and the calculation of the vertical reference separation adaptive interpolation filtering performed to obtain the second interpolation image. The complexity can be calculated separately, and archive filtering with low computational complexity can be determined as the final interpolation filtering method.

The determination unit 173 generates interpolation filtering information based on the determined final interpolation filtering method (S730). Here, the interpolation filtering information includes at least one of an interpolation filtering indicator indicating a type of interpolation filtering applied to the final interpolation filtering method and a filter coefficient of interpolation filtering applied to the final interpolation filtering method.

Subsequently, the entropy encoder 140 entropy-codes the generated interpolation filtering information, inserts it into a bitstream, and transmits the encoded interpolation filtering information to the decoding apparatus 400 (S740).

8 is a block diagram showing a configuration of a decoding apparatus according to an embodiment of the present invention.

Referring to FIG. 8, the decoding apparatus 800 according to the embodiment of the present invention includes an entropy decoding unit 810, an inverse quantization and inverse DCT unit 820, a motion compensator 830, an adder 840, A deblocking unit 840 and an interpolation filter 860.

The entropy decoder 810 entropy decodes the bit stream to extract transform coefficients, motion vectors, and interpolation filtering information. That is, the entropy decoding unit 810 entropy decodes the encoded interpolation filtering information received from the encoding apparatus 100.

The inverse quantization and inverse DCT unit 820 inversely quantizes the extracted transform coefficient and inverse discrete cosine transform to obtain a differential block.

The motion compensator 830 generates the prediction block of the current block by using the motion vector extracted by the entropy decoder 810.

The adder 840 reconstructs the current block by adding the difference block obtained by the inverse quantization and inverse DCT unit 820 and the prediction block generated by the motion compensation unit 830.

The deblocking unit 850 deblocks and outputs the current block restored by the adder 840.

The interpolation filter 860 interpolates the reference image using the interpolation filtering information extracted by the entropy decoder 810 to obtain an interpolation image.

Here, the interpolation filtering information includes at least one of an interpolation filtering indicator indicating the type of interpolation filtering applied to the final interpolation filtering method determined by the encoding apparatus 100, and filter coefficients of the interpolation filtering applied to the final interpolation filtering method.

That is, the interpolation filter 860 may know interpolation filtering applied to an image encoded by the encoding apparatus 100 based on interpolation filtering information. In this way, the interpolation filter 860 interpolates and filters the image encoded by the interpolation filtering applied by the encoding apparatus 100.

In addition, the interpolation filter 860 decodes an image encoded and transmitted by the encoding apparatus 100 based on the obtained interpolated image to obtain a reconstructed image. Here, since the content of decoding the encoded image to obtain the reconstructed image is well known, a detailed description thereof will be omitted.

9 is a flowchart provided to explain a decoding method using interpolation filter information according to an embodiment of the present invention.

First, the entropy decoder 810 receives interpolation filtering information from the encoding apparatus 100 (S910).

In detail, the entropy decoding unit 810 extracts interpolation filtering information from the bit stream received from the encoding apparatus 100, and entropy decodes the extracted interpolation filtering information.

Here, the interpolation filtering information includes at least one of an interpolation filtering indicator indicating the type of interpolation filtering applied to the final interpolation filtering method determined by the encoding apparatus 100, and filter coefficients of the interpolation filtering applied to the final interpolation filtering method.

Subsequently, the interpolation filter 860 determines an interpolation filtering method used in encoding based on the interpolation filtering information (S920). That is, the interpolation filter 860 determines the final interpolation filtering method determined by the encoding apparatus 100 based on the interpolation filtering information.

The interpolation filter 860 obtains an interpolation image based on the determined interpolation filtering method (S930). That is, the interpolation filter 860 interpolates and filters the reference image as a final interpolation filtering method to obtain an interpolated image.

Next, the interpolation filter 860 decodes an encoded image based on the obtained interpolated image to obtain a reconstructed image (S940).

1 to 4, the interpolation filtering information has been described as interpolation filtering indicators and filter coefficients of interpolation filtering. However, interpolation filtering information includes context information of the interpolation filtering applied to the final interpolation filtering method, and reconstructed image and current. It may include at least one of statistical characteristics of the residual signal of one or more of the images, and distribution of pixel values. Here, the statistical characteristic of the residual signal includes at least one of discrete cosine (Ccrete) value of the residual signal and coded block pattern (CBP), and an example of the residual signal may be a differential block.

In this case, in the interpolation filter 170 illustrated in FIG. 2, the order of the interpolation unit 171 and the determination unit 173 may be changed. In other words, before the interpolator 171 acquires the first interpolated image and the second interpolated image, the determiner 173 uses the context information of the interpolation filtering, the statistical characteristics of the residual signal, or the distribution of pixel values. Interpolation filtering can be determined. Through this, the interpolator 171 may obtain the interpolated image by interpolating the reference image by the determined final interpolation filtering method.

Accordingly, the decoding apparatus 800 may infer and obtain interpolation filtering information from already decoded images without receiving interpolation filtering information from the encoding apparatus 200. That is, step S910 may be omitted in FIG. 9, and interpolation filtering information may be inferred from the images decoded in step S920. Here, since the information of inferring the interpolation filtering information from the decoded images is a well known technique, a detailed description thereof will be omitted.

In the above description, the final interpolation filtering method is determined by selecting one of the first interpolation filtering and the second interpolation filtering, but the present invention is not limited thereto. Rather, one of three or more interpolation filtering may be selected to determine a final interpolation filtering method.

In the above description, the image is divided into blocks and the present invention can be implemented in blocks. That is, each block may select and use an interpolation filter method identically or differently.

In the above description, the current image is predicted using the obtained first and second interpolation images, but the present invention is not limited thereto, and the current image may be predicted by using the first interpolation image and the second interpolation image together. .

That is, the determiner 173 may generate an average interpolation image that is an average of the obtained first interpolation image and the second interpolation image, and predict the current image based on the generated average interpolation image.

In addition, the determiner 173 may predict the current image by using the weight. That is, the determiner 173 may generate an average interpolation image by weighting each of the first and second interpolation images, and predict the current image based on the generated average interpolation image.

Similarly, the decoding apparatus 800 may predict the current image by using the first interpolation image and the second interpolation image, and decode the encoded image by using the predicted current image.

As described above, the present invention has been described by way of limited embodiments and drawings, but the present invention is not limited to the above embodiments, and those skilled in the art to which the present invention pertains various modifications and variations from such descriptions. This is possible.

Therefore, the scope of the present invention should not be limited to the described embodiments, but should be determined not only by the claims below but also by the equivalents of the claims.

1 is a block diagram showing a configuration of an encoding apparatus according to an embodiment of the present invention.

2 is a block diagram showing the configuration of an interpolation filter according to an embodiment of the present invention.

3 and 4 are diagrams provided to explain horizontal reference adaptive interpolation filtering according to an embodiment of the present invention.

5 and 6 are diagrams provided to explain vertically adaptive interpolation filtering according to an embodiment of the present invention.

7 is a flowchart provided to explain an interpolation filtering method according to an embodiment of the present invention.

8 is a block diagram showing a configuration of a decoding apparatus according to an embodiment of the present invention.

9 is a flowchart provided to explain a decoding method using interpolation filter information according to an embodiment of the present invention.

<Explanation of symbols for the main parts of the drawings>

110: motion estimation unit

120: motion compensation unit

125, 155: first difference part, second difference part

130: DCT and quantization unit

140: entropy encoder

150: dequantization and inverse DCT

160: deblocking unit

170: interpolation filter

Claims (9)

Obtaining a plurality of interpolated images by filtering the reference images with a plurality of interpolation filtering; And Determining any one of the plurality of interpolation filtering methods by using the plurality of interpolation images as a final interpolation filtering method Interpolation filtering method comprising a. The method of claim 1, The determining step, And interpolation filtering applied to the interpolation image when the encoding rate is high when the current image is predicted based on the obtained plurality of interpolation images, by using the final interpolation filtering method. The method of claim 1, The determining step, An interpolation filtering method having the lowest computational complexity for performing interpolation filtering in each of the plurality of interpolation filterings as the final interpolation filtering method. The method of claim 1, The plurality of interpolation filtering includes at least one of separate adaptive interpolation filtering, directional adaptive interpolation filtering, adaptive interpolation filtering, interpolation filtering using a fixed filter coefficient, The splitting adaptive interpolation filtering includes at least one of vertical reference splitting adaptive interpolation filtering and horizontal reference splitting adaptive interpolation filtering. The method of claim 1, Storing interpolation filtering information indicating the determined final interpolation filtering method Interpolation filtering method further comprising. Vertical adaptive interpolation filtering of the reference image; And Acquiring an interpolation image by performing vertical adaptive interpolation filtering on the reference image after the vertical adaptive interpolation filtering Interpolation filtering method comprising a. Determining an interpolation filtering method used in encoding based on the interpolation filtering information; Obtaining an interpolation image based on the determined interpolation filtering method; And Obtaining a reconstructed image by decoding the encoded image based on the interpolated image. Decryption method comprising a. The method of claim 7, wherein Receiving the interpolation filtering information More, The determining step, Determine the interpolation filtering method based on the received interpolation filtering information, The received interpolation filtering information includes at least one of an interpolation filtering indicator indicating a type of interpolation filtering applied to a final interpolation filtering method used in encoding, and a filter coefficient. The method of claim 7, wherein Inferring the Archive Filtering Information from Decoded Images More, The determining step, Determine the interpolation filtering method based on the inferred interpolation filtering information, The inferred interpolation filtering information includes at least one of context information, statistical characteristics of a residual signal of at least one of a reconstructed image and current images, and a distribution of pixel values.

Images