KR101934840B1 - Method and apparatus for image interpolation having quarter pixel accuracy using intra prediction modes - Google Patents

Abstract

The present invention relates to an image interpolation technique of 1/4 pixel resolution used in motion estimation and motion compensation of 1/4 pixel resolution. An image interpolation method having a quarter pixel resolution using an intra mode includes the steps of obtaining an optimal intra mode from among a plurality of intra modes, interpolating an input image with half-pixel accuracy, Classifying the quotient pixel position into a second classification and otherwise classifying the quotient pixel position into a first classification if all surrounding pixels for a quarter pixel position are already restored before interpolation with pixel precision; The method comprising the steps of: interpolating a quota pixel belonging to a quarter pixel position belonging to a first classification, irrespective of the optimal intra mode; comparing a quarter pixel belonging to a quarter pixel position belonging to the second classification with a direction indicated by the optimal intra mode And interpolating.

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method and an apparatus for interpolating an image having a quarter pixel resolution using an intra mode,

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a video compression method, and more particularly, to a video interpolation technique of 1/4 pixel resolution used in 1/4 pixel resolution motion estimation and motion compensation.

As information and communication technologies including the Internet are developed, not only text and voice but also video communication are increasing. Conventional character - oriented communication methods are not sufficient to satisfy various needs of consumers. Accordingly, multimedia services capable of accommodating various types of information such as text, images, and music are increasing. The amount of multimedia data is so large that it needs a large capacity storage medium and requires a wide bandwidth in transmission. Therefore, it is necessary to use a compression coding technique to transmit multimedia data including characters, images, and audio.

The basic principle of compressing data is the process of eliminating the redundancy of data. A temporal redundancy such as a spatial redundancy such that the same color or object is repeated in an image, a temporal redundancy such as a case where the adjacent frame is almost unchanged in the moving picture frame or the same sound continuously repeats in the audio, The data can be compressed by eliminating the psychological visual overlap considering the insensitivity to the data.

Various video coding standards such as Moving Picture Experts Group (MPEG) -2, MPEG-4, and H.264 have appeared in order to standardize the moving image compression technique. As shown in FIG. 1, all video coding techniques employ a technique called block motion estimation to eliminate temporal redundancy between adjacent video frames.

For example, to encode a block 12 in the current frame 10, a block 17 matching the block 12 in a reference frame 15 at a different temporal location from the current frame 10 Find. Thereafter, the residual between the block 12 of the current frame 10 and the block 17 of the reference frame 15 is obtained, and the difference is encoded to improve the coding efficiency. Here, the displacement between the blocks is represented as a motion vector, and motion compensation is performed on the reference frame 15 by the motion vector.

By using the motion compensated frame as a prediction image and subtracting it from the original image, the size of data to be coded can be reduced.

FIG. 2 is a diagram for explaining a method of interpolating an image at 1/4 pixel resolution according to the conventional H.264. In Fig. 2, among the total of 16 pixels at the positions of x and y, a gray pixel indicates an integer pixel, a red pixel indicates a half pixel, and a yellow or green pixel indicates a quarter pixel.

For improved coding efficiency, H.264 motion estimation and compensation is based on a quarter pixel resolution as described above. Therefore, in order to perform the motion compensation prediction, the reference image must be interpolated at a 1/4 pixel resolution. In this process, a 6-tap Wiener interpolation filter is used in H.264.

2, B (x, y-1), B (x, y), and B (x, y-2) (x, y + 1), B (x, y + 2), B ), H (x + 2, y) and H (x + 3, y) are calculated using a 6-tap Winner filter. Then, J (x, y) is transformed into B (x, y-1), B (x, y) , B (x, y + 2), and B (x, y + 3).

A (x, y), C (x, y) and I (x, y) and K (x, y) are interpolated horizontally using samples of adjacent integer pixels or halfpixel positions. Then, D (x, y), F (x, y), L (x, y) and N (x, y) are interpolated vertically.

Finally, E (x, y), G (x, y), M (x, y) and O (x, y) are interpolated diagonally.

Because this simple interpolation process is used in H.264, the computational complexity can be significantly reduced. However, since the regional characteristics of the reference image are not considered, the prediction efficiency is inevitably reduced.

Therefore, the present invention proposes an adaptive interpolation technique based on information from 4x4 intra prediction. Thus, the present invention is based on the assumption that the optimal intra mode is likely to have a very high correlation with the directionality between the pixels.

It is an object of the present invention to provide a method and apparatus for motion interpolation with 1/4 pixel resolution that have higher accuracy and do not significantly increase the amount of computation.

The technical objects of the present invention are not limited to the technical matters mentioned above, and other technical subjects not mentioned can be clearly understood by those skilled in the art from the following description.

According to another aspect of the present invention, there is provided a method of interpolating an image having a quarter pixel resolution using an intra mode, comprising the steps of: (a) obtaining an optimal intra mode from among a plurality of intra modes; (b) interpolating the input image with half pixel precision; (c) if all neighboring pixels for a quarter pixel position have already been reconstructed before interpolation from the interpolated half-pixel accuracy image at quarter pixel precision, the quotient pixel position is assigned as a second classification, Classification into a first classification; (d) interpolating a quarter pixel belonging to a quarter pixel position belonging to the first classification, irrespective of the optimal intra mode; And (e) interpolating a quarter pixel belonging to a quarter pixel position belonging to the second classification with reference to a direction indicated by the optimal intra mode.

According to an aspect of the present invention, there is provided an apparatus for interpolating an image having a quarter pixel resolution using an intra mode, the apparatus comprising: an intra predictor for obtaining an optimal intra mode among a plurality of intra modes; A half pixel interpolator for interpolating the input image with half pixel precision; If all the surrounding pixels for a quarter pixel position have already been reconstructed before interpolation from the interpolated image at the halfpixel accuracy to the quarter pixel precision, the quotient pixel position is set to the second classification, otherwise, A quarter-pixel classification unit for classifying the pixels into a plurality of pixels; A first interpolation unit interpolating a quarter pixel belonging to a quarter pixel position belonging to the first classification, irrespective of the optimal intra mode; And a second interpolator interpolating a quarter pixel belonging to a quarter pixel position belonging to the second classification with reference to a direction indicated by the optimal intra mode.

According to the present invention, as compared with the conventional motion interpolation technique of quarter pixel resolution in H.264, since the already obtained intra mode is used as it is, Motion interpolation becomes possible.

Brief Description of the Drawings Fig. 1 is a diagram showing the concept of normal block motion estimation; Fig.
BACKGROUND OF THE INVENTION 1. Field of the Invention [0001]
3 is a block diagram illustrating an image interpolating apparatus with quarter pixel resolution according to an embodiment of the present invention.
4 is a diagram showing nine intra modes of H.264;
5 is a view showing a quarter pixel belonging to the second classification and surrounding pixels.
6 is a block diagram showing the configuration of a video encoder according to an embodiment of the present invention;
7 is a block diagram showing the configuration of a video decoder according to an embodiment of the present invention;

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS The advantages and features of the present invention, and the manner of achieving them, will be apparent from and elucidated with reference to the embodiments described hereinafter in conjunction with the accompanying drawings. The present invention may, however, be embodied in many different forms and should not be construed as being limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art. Is provided to fully convey the scope of the invention to those skilled in the art, and the invention is only defined by the scope of the claims. Like reference numerals refer to like elements throughout the specification.

3 is a block diagram illustrating a video interpolation apparatus 100 of quarter pixel resolution according to an embodiment of the present invention. The image interpolating apparatus 100 includes an intra predictor 105, a half pixel interpolator 110, a quarter pixel classifier 120, a first interpolator 130, a second interpolator 140, and a pixel buffer 150 ). ≪ / RTI >

The intra prediction unit 105 predicts an optimal directional intra mode for the current block (hereinafter referred to as an intra mode). Specifically, the optimal intra mode can be computed based on a cost function C such as Equation 1 based on a rate-distortion. Here, E denotes the difference between the reconstructed signal and the original signal by decoding the encoded bit, and B denotes a bit amount required in each coding. Also, λ denotes a coefficient capable of adjusting the reflection ratio of E and B as a Lagrangian coefficient.

[Equation 1]

C = E + lambda x B

In H.264, there are a total of nine modes as shown in FIG. 4 in the intra mode. That is, prediction is performed in the various intra modes from the pixels (maximum 13 pixels of A to M) adjacent to the current block to obtain the cost function, and then the intra mode in which the cost function is minimized is referred to as the intra mode . The determined intra mode indicates that the correlation among the pixels in the corresponding direction is the highest. For example, in the case of mode 0, it means that the correlation among the pixels in the vertical direction is the highest.

In this case, if the video encoder performs image interpolation, it is necessary to calculate the optimal intra mode. However, if the video decoder performs image interpolation, a separate intra mode calculation is unnecessary, and intra mode information .

The half-pixel interpolator 110 interpolates an input image having an integer pixel resolution into an image having a half-pixel resolution. This interpolation is a process of interpolating a red half pixel from a gray integer pixel in Fig. 2, and it is also possible to use the same 6-tap Winner filter as in the conventional H.264.

The quarter pixel classification unit 120 classifies the pixel into a quarter pixel belonging to the first classification and a quarter pixel belonging to the second classification according to the type (position) of the quarter pixels before performing the interpolation at the quarter pixel resolution. A quarter pixel position belonging to the first classification is provided to the first interpolator 130 and a quarter pixel position belonging to the second classification is provided to the second interpolator 140. [

The first classification indicates a case where a pixel that is not restored among surrounding pixels (for example, eight pixels) surrounding the corresponding quarter pixel exists, and the second classification indicates that all the surrounding pixels surrounding the corresponding quarter pixel are restored Respectively.

The first interpolator 130 interpolates the quarter pixels belonging to the first classification according to the conventional method, referring to the quarter pixel positions belonging to the first classification. 2, pixels A (x, y), C (x, y), and D (x, y) are displayed in yellow among the quarter pixels (pixels indicated by yellow or green as described above) (X, y), F (x, y), I (x, y), K It is not easy to apply the intra mode proposed in the present invention. Therefore, for these pixels, for example, interpolation is performed in the same manner as in the conventional H.264.

(X, y) and K (x, y) are horizontally interpolated using samples of adjacent integer pixels or half pixel positions, and D (x, y) y), F (x, y), L (x, y) and N (x, y) are interpolated vertically using samples of adjacent integer pixels or halfpixel positions.

Meanwhile, the second interpolator 140 refers to the quarter pixel position belonging to the second classification, and interpolates the quarter pixel belonging to the second classification in consideration of the directionality according to the calculated intra mode. For example, E (x, y), G (x, y), M (x, y) and O (x, y), which are green pixels in FIG. 2, Therefore, it is possible to perform interpolation considering directionality. The position of a quarter pixel belonging to the second classification can also be regarded as a position of surrounding pixels at four corner positions of the integer pixel when viewed from one integer pixel.

However, in the case of a video encoder, since the optimum intra mode is calculated irrespective of whether a particular pixel belongs to an inter macroblock or an intra macroblock, the calculated information is always available, but in the case of a video decoder, It is needless to say that interpolation considering the directionality is possible only when belonging to a macroblock.

Specifically, a specific method of interpolating the quarter pixel belonging to the second classification will be described with reference to FIG.

In FIG. 5, a total of eight pixels surround the E (x, y) among the quarter pixels to be interpolated. In this case, if the intra mode of the integer pixel at the position of x, y, that is, the intra mode of P (x, y) is expressed as P_mode (x, y) Can be interpolated by the same pseudo code.

If (P_mode (x, y) = 1, 6 or 8)
E (x, y) = [D (x, y) + F
Else If (P_mode (x, y) = 0, 5 or 7)
E (x, y) = [A (x, y) + I (x, y)] / 2
Else If (P_mode (x, y) = 4)
E (x, y) = [P (x, y) + J
Else If (P_mode (x, y) = 3)
E (x, y) = [B (x, y) + H
Else
(X, y) + I (x, y) + I (x, y)

That is, when the intra mode is 1 (horizontal), 6 (horizontal down), or 8 (horizontal up), since the association is high in the horizontal direction between the pixels, Interpolates the pixels.

When the intra mode is 0 (vertical), 5 (vertical right), or 7 (vertical left), since the correlation is high in the vertical direction between the pixels, the current quarter pixel is interpolated by the average of the upper and lower pixels among the surrounding pixels .

When the intra mode is 4 (diagonal down-right), since the correlation between the pixels is high in the lower right diagonal direction, the current quarter pixel is interpolated by the average between the upper left pixel and the lower right pixel among the surrounding pixels.

When the intra mode is 3 (diagonal down-left), since the correlation is high in the horizontal direction between the pixels, the current quarter pixel is interpolated by the average between the upper right pixel and the lower left pixel among the surrounding pixels.

In other cases, that is, when the intra mode is 2 (DC), since there is no directionality between the pixels, the current quarter pixel is interpolated by the average of the four pixels closest to the current quarter pixel.

The interpolation of E (x, y) has been described above, but the rest G (x, y), M (x, y) and O Of course it is.

Finally, the pixel buffer 150 is composed of an integer pixel obtained from the original input image, a half pixel obtained from the half pixel interpolator 110, and a combination of the first interpolator 130 and the quarter pixel obtained from the second interpolator 140 Thereby generating an interpolated image having a quarter pixel resolution.

6 and 7 are diagrams for explaining a practical environment to which a quarter pixel interpolation method according to the present invention can be applied. 6 is a block diagram of a video encoder, and Fig. 7 is a block diagram of a video decoder.

6 is a block diagram illustrating a configuration of a video encoder 200 according to an embodiment of the present invention.

The video encoder 200 includes an intra prediction unit 210, a spatial transform unit 220, a quantization unit 230, an entropy coding unit 240, a motion estimation unit 250, a motion compensation unit 260, An inverse quantization unit 271, an inverse spatial transform unit 272, an intra reconstruction unit 273 and an image interpolation apparatus 100 according to an embodiment of the present invention.

The selecting unit 280 selects an advantageous prediction method from the intra prediction and the inter prediction. Normally, this selection is made in macroblock units. The selecting unit 280 performs the actual coding on the two prediction methods and selects a prediction method having a lower cost. Here, the cost C can be defined in various ways, and can be typically defined as a cost function based on the above-described rate-distortion.

The intraprediction unit 210 determines an optimal intra mode for a current block (e.g., 4x4 block) among nine intra modes, as a first step. The determination of this intra mode can also be determined in terms of the rate-distortion cost described above. The intra predictor 210 generates a predictive block according to the determined intra mode, and obtains a residual block obtained by subtracting the current block from the predictive block.

The residual block is provided to the spatial transformer 220 to perform spatial transform such as DCT.

Meanwhile, the image interpolating apparatus 100 may be used in an operation related to inter prediction that is used complementarily with the intra prediction.

The motion estimation unit 250 performs motion estimation of the current frame and obtains a motion vector from among the input video frames with reference to the reference frame. A widely used algorithm for such motion estimation is a block matching algorithm. However, in the standard codec such as H.264, the motion estimation of the quarter pixel resolution is performed, so that the search area of the current frame and the search frame of the reference frame must be interpolated at quarter pixel resolution, respectively. The image interpolating apparatus 100 performs interpolation of this quarter pixel resolution in the same manner as in the embodiment of Fig. Of course, when the video encoder 200 is provided with a separate intra prediction unit 210 as shown in FIG. 6, the intra prediction unit 105 of FIG. 3 may be omitted. The motion estimator 250 provides the entropy encoder 240 with motion data such as a motion vector, a size of a motion block, and a reference frame number, which are obtained as a result of motion estimation.

The motion compensation unit 260 reduces the temporal redundancy of the input video frame. In this case, the motion compensating unit 260 performs motion compensation on the reference frame using the motion vector calculated by the motion estimating unit 250, thereby generating an inter prediction frame for the current frame. H.264 also requires interpolation of quarter pixel resolution in the motion compensation process. Accordingly, the image interpolating apparatus 100 can perform the interpolation of the quarter pixel resolution in the motion compensation process as well. Of course, if motion estimation and compensation are performed in a single process (that is, a method in which an interpolated image is stored in a memory in motion estimation and read and used in motion compensation), a separate interpolation may be unnecessary in motion compensation have.

The difference 215 eliminates the temporal redundancy of the video by subtracting the current frame from the inter prediction frame. The output from the difference branch 215, that is, the residual block in the inter prediction is provided to the spatial transform unit 220. [

The spatial transform unit 220 removes the spatial redundancy using the spatial transform method on the residual block provided from the intra prediction unit 210 or the difference filter 215. [ DCT (Discrete Cosine Transform) is mainly used for the spatial transformation. Coefficients obtained by spatial transformation are called transformation coefficients.

The quantization unit 230 quantizes the transform coefficients obtained by the spatial transform unit 220. The term 'quantization' refers to an operation of dividing the transformation coefficient expressed by an arbitrary real number value into a predetermined section, representing the transformation coefficient as a discrete value, and matching it with a predetermined index.

The entropy encoding unit 240 losslessly encodes the transform coefficients quantized by the quantization unit 230, the motion data provided by the motion estimation unit 250 or the intra mode provided from the intra prediction unit 210, . As the lossless coding method, arithmetic coding, variable length coding, Huffman coding, or the like can be used.

On the other hand, the video encoder 200 uses a closed-loop video encoding technique in order to reduce a drifting error between the encoder stage and the decoder stage. To this end, it is necessary to further include an inverse quantization unit 271, an inverse spatial transform unit 272, an intra reconstruction unit 273, and the like.

The inverse quantization unit 271 inversely quantizes the quantized coefficients in the quantization unit 230. [ This dequantization process corresponds to the inverse of the quantization process. The inverse spatial transform unit 272 inversely transforms the inverse quantized result and provides it to the adder 225 or the intra restoration unit 273. [ In this case, the residual image reconstructed by the inverse spatial transform is provided to the intra reconstruction unit 273 if it is a frame originally generated by intra prediction, and is provided to the adder 225 if the image is generated by inter prediction.

The adder 225 restores the video frame by adding the residual image provided from the inverse spatial transformer 272 and the previous frame stored in the frame buffer (not shown) provided from the motion compensator 260, And provides the video frame to the motion estimation unit 250 as a reference frame.

The intra restoration unit 273 restores the current block from the prediction blocks obtained from neighboring blocks by using the residual block constituting the residual image and the intra mode value. The current block can be restored by simply summing the residual block and the prediction block, and the value of the intra mode is used to obtain the prediction block. The process of obtaining the prediction block based on the determined intra mode is the same as in the intra prediction unit 210. [

7 is a block diagram illustrating the configuration of a video decoder 500 according to an embodiment of the present invention.

The video decoder 500 includes an entropy decoding unit 510, an inverse quantization unit 520, an inverse spatial transform unit 530, an intra restoration unit 540, a motion compensation unit 550, and an image interpolation device 100 .

The entropy decoding unit 510 performs lossless decoding in the reverse of the entropy coding system to extract motion data (motion vector and reference frame, etc.) in inter prediction, intra mode and texture data in intra prediction. The texture data is supplied to the inverse quantization unit 520, the motion data is supplied to the motion compensation unit 550, and the intra mode is provided to the intra restoration unit 540.

The inverse quantization unit 520 inversely quantizes the texture data transmitted from the entropy decoding unit 510. The inverse quantization process is a process of expressing a predetermined index at the end of the encoder 200 and searching for a quantized coefficient matched with the value.

The inverse spatial transform unit 530 performs inverse spatial transform and transforms the coefficients (frequency domain) generated as a result of the inverse quantization into residual blocks in the spatial domain. For example, if the video encoder unit is spatially transformed in the DCT scheme, an inverse DCT transform will be performed.

The intra-reconstructing unit 540 reconstructs the current block from the prediction block obtained from the neighboring blocks already generated using the intra-mode value transmitted from the entropy decoding unit 510 and the residual block provided from the inverse spatial transform unit 530, Restore the block. The current block can be restored by simply summing the residual block and the prediction block, and the value of the intra mode is used to obtain the prediction block. The process of obtaining the prediction block based on the determined intra mode value is the same as that of the intra prediction unit 210 in the video encoder 200.

On the other hand, for restoration from inter prediction, a motion compensation unit 550 is used. The motion compensating unit 550 generates motion compensated frames by performing motion compensation on the previously reconstructed frames using the motion data provided from the entropy decoding unit 510.

However, since a standard codec such as H.264 performs motion estimation with a quarter pixel resolution, it is necessary to interpolate a quarter pixel resolution in such motion compensation. The image interpolating apparatus 100 performs interpolation of this quarter pixel resolution in the same manner as in the embodiment of Fig. Of course, in the case of the video decoder 500, a separate intra predictor (105 in FIG. 3) is unnecessary. When only the information about the intra mode provided from the entropy decoding unit 510 is provided to the image interpolator 100 do.

The adder 515 restores the current block by adding the residual block and the motion compensated frame provided from the motion compensating unit 550 when the residual block reconstructed in the inverse spatial transform unit is generated by the inter prediction.

As a result, a restored frame can be generated by combining the restored block in the intra restoration unit 540 and the restored block in the adder 515.

Up to this point, each of the components of FIGS. 3, 6, and 7 may refer to software or hardware such as a field-programmable gate array (FPGA) or an application-specific integrated circuit (ASIC). However, the components are not limited to software or hardware, and may be configured to be in an addressable storage medium and configured to execute one or more processors. The functions provided in the components may be implemented by a more detailed component or may be implemented by a single component that performs a specific function by combining a plurality of components.

To simulate the interpolation algorithm according to the present invention, a JM 12.4 encoder was used and tested on CIF sized video sequences (Akiyo, Bus, Paris, Mobile, Stefan, etc.). In order to evaluate the rate distortion performance of an interpolation algorithm according to the present invention, the present invention is compared to H.264. Here, the motion search range was selected to be 32, and the number of reference frames was selected to be 1. In addition, rate distortion optimization and context-adaptive variable length coding (CAVLC) are applied, and the structure of group of pictures (GOP) is IPPP. The CIF sequences were tested for quantization parameters (QP) of 28, 32 and 36, respectively. In Table 2, ΔPSNR and ΔBitrate represent changes in PSNR and bit rate when compared with H.264, respectively.

QP Sequence ? PSNR (dB) ? Bitrate (%) 28 Bus 0.01 -0.04 Paris 0.01 -0.14 Mobile -0.01 -0.91 Stefan 0.01 -0.31 32 Bus 0 0.06 Paris 0 -0.19 Mobile -0.01 -1.61 Stefan 0 -0.18 36 Bus -0.01 0.25 Paris -0.01 -0.08 Mobile 0 -1.41 Stefan 0.01 0.31 Average 0 -0.354

From the results shown in Table 2, it can be seen that according to the present invention, the PSNR does not change on average in comparison with the existing H.264, while the bit rate is reduced by 0.354% on average. The reduction of the bit rate may not be very large from the viewpoint of the entire bit stream, but it can be seen that the reduction effect is considerable only considering the interpolation process of the quarter pixel. Particularly, the interpolation from a single integer pixel to a total of 15 subpixels shows that the effect is significant considering that the modified method is applied to only four quarter pixels.

While the present invention has been described in connection with what is presently considered to be practical exemplary embodiments, it is to be understood that the invention is not limited to the disclosed embodiments, but, on the contrary, You will understand. It is therefore to be understood that the embodiments described above are in all respects illustrative and not restrictive.

100: image interpolating apparatus 105: intra prediction unit
110: Half pixel interpolator 120: Quarter pixel classification unit
130: First interpolation part 140: Second interpolation part
150: Pixel buffer

Claims (12)

A video decoding method of a video decoder,
Arithmetically decoding the bit stream to obtain a motion vector, a transform coefficient, and a reference image number;
Wherein the inter prediction block is obtained by using the reference image and the motion vector, the inter prediction block is obtained by an interpolation pixel obtained by the reference image, and the one of the interpolation pixels is a first And a different interpolation method is applied to the quarter pixel belonging to the first classification and the quarter pixel belonging to the second classification;
Transforming the transform coefficients to obtain a residual block; And
And adding the residual block and the inter prediction block to reconstruct a current block,
Wherein a quotient pixel belonging to the first classification is obtained by using an integer pixel,
The quota pixel belonging to the second classification is obtained by using a quarter pixel
/ RTI > The method according to claim 1,
The quater pixel belonging to the second classification is obtained by using at least one of pixels in the vertical direction, pixels in the horizontal direction, pixels in the lower right diagonal direction, and pixels in the lower left diagonal direction
/ RTI > 3. The method of claim 2,
In the case where the intra mode is the vertical mode, the quota pixel belonging to the second classification is obtained by using pixels in the vertical direction
/ RTI > 3. The method of claim 2,
In the case where the intra mode is the horizontal mode, the quota pixel belonging to the second classification is obtained by using pixels in the horizontal direction
/ RTI > The method according to claim 1,
Wherein the reference image comprises a first integer pixel, a second integer pixel located to the right of the first integer pixel, a third integer pixel located below the first integer pixel, and a third integer pixel located to the right of the third integer pixel A fourth integer pixel,
The right quater pixel of the first integer pixel, the left quater pixel of the second integer pixel, the lower quater pixel of the first integer pixel, and the upper quater pixel of the third integer pixel belong to the first classification,
The lower right quater pixel of the second integer pixel, the upper right quater pixel of the third integer pixel, and the upper left quater pixel of the fourth integer pixel correspond to the second classification under
/ RTI > The method according to claim 1,
Wherein a quota pixel belonging to the second classification is obtained by using a quota pixel belonging to the first classification
/ RTI > The method according to claim 6,
Wherein a quota pixel belonging to the second classification is obtained by using a vertical quater pixel belonging to the first classification
/ RTI > delete delete A video decoder device comprising:
A decoding unit for arithmetically decoding the bit stream to obtain a motion vector, a transform coefficient, and a reference image number;
Wherein the inter prediction block is obtained by using the reference image and the motion vector, the inter prediction block is obtained by an interpolation pixel obtained by the reference image, and the one of the interpolation pixels is a first A motion compensator which is classified into a classifier and a second classifier and to which a different interpolation scheme is applied to a quarter pixel belonging to the first classifier and a quotient pixel belonging to the second classifier;
An inverse spatial transformer for transforming the transform coefficients to obtain a residual block; And
And an adder for adding the residual block and the inter prediction block to reconstruct a current block,
Wherein a quotient pixel belonging to the first classification is obtained by using an integer pixel,
The quota pixel belonging to the second classification is obtained by using a quarter pixel
A video decoding apparatus. delete A video encoding method of a video encoder,
Obtaining a motion vector;
Wherein the inter prediction block is obtained by an interpolation pixel obtained by the reference image using a reference image and the inter prediction block for the current block is obtained using the reference image, And a second interpolation method is applied to a quarter pixel belonging to the first classification and a quarter pixel belonging to the second classification;
Obtaining a residual block by subtracting the inter prediction block from the current block;
Obtaining a transform coefficient using the residual block;
And generating a bitstream by arithmetically coding the transform coefficient, the motion vector, and the number of the reference image,
Wherein a quotient pixel belonging to the first classification is obtained by using an integer pixel,
The quota pixel belonging to the second classification is obtained by using a quarter pixel
Video encoding method.

Images