WO2014084674A2 - Intra prediction method and intra prediction apparatus using residual transform - Google Patents

Abstract

The present invention relates to a method and apparatus for performing intra prediction. The method comprises the steps of: generating a residual signal according to an intra prediction mode; performing a residual transform by applying a pixel-based DPCM in a first direction to the residual signal; and performing cross residual transform by applying the pixel-based DPCM in a second direction to the residual-transformed residual signal, wherein the residual transform and cross residual transform are performed on a selected region on the basis of a specific object or a block.

Description

Intra prediction method and apparatus using residual transformation

The present invention relates to a method and apparatus for image encoding / decoding, and more particularly, to a method of performing intra prediction by using a residual transform during encoding and decoding.

In general, in video coding, a residual signal is generated using intra prediction and inter prediction. The reason for obtaining the residual signal is that when coding with the residual signal, the amount of data is small and the data compression ratio is high, and the better the prediction, the smaller the value of the residual signal.

The intra prediction method predicts data of the current block by using pixels around the current block. The difference between the actual value and the predicted value is called the residual signal block. In the case of HEVC, the intra prediction method is increased from 35 prediction modes as shown in FIG. 1 to 9 prediction modes used in the existing H.264 / AVC to further refine the prediction (however, the planar prediction mode and the DC prediction mode are shown in FIG. Invisible from 1).

In the case of the inter prediction method, the most similar block is found by comparing the current block with blocks in neighboring pictures. At this time, the position information (Vx, Vy) of the found block is called a motion vector. The difference between pixel values in a block between the current block and the prediction block predicted by the motion vector is called a residual signal block (motion-compensated residual block).

As such, intra prediction and inter prediction are further subdivided to reduce the amount of data of the residual signal, and a video encoding and decoding method having a small amount of computation is required without degrading codec performance by using an efficient transform.

An embodiment of the present invention provides a video encoding and decoding method having a good performance with a small operation amount in a transform coding process of a video codec, and an apparatus therefor.

However, the technical problem to be achieved by the embodiment of the present invention is not limited to the technical problems as described above, and other technical problems may exist.

As an technical means for achieving the above technical problem, the intra prediction method according to an embodiment of the present invention generates a residual signal according to the intra prediction mode, by applying a pixel unit DPCM of the first direction to the residual signal Performing a residual transform and performing a cross residual transform by applying a pixel unit DPCM in a second direction to the residual transformed residual signal, wherein the residual transform and the cross residual transform are based on a specific object or block Is performed for the selected area.

In addition, the intra prediction apparatus according to the embodiment of the present invention, a residual signal generation unit for generating a residual signal according to the intra prediction mode, a residual for performing the residual conversion by applying a pixel unit DPCM of the first direction to the residual signal A transform performing unit and a cross residual transform performing unit performing cross residual transform by applying a pixel unit DPCM in a second direction to the residual transformed residual signal, wherein the residual transform and cross residual transform are based on a specific object or block Is performed for the selected area.

Meanwhile, the intra prediction method may be implemented as a computer-readable recording medium that records a program to be executed in a computer.

According to an embodiment of the present invention, by selectively performing residual transform and crossed residual transform on the selected region instead of the existing intra prediction, the magnitude of the residual signal for the selected region is minimized. Compression performance can be improved.

1 is a diagram illustrating examples for intra prediction modes.

2 is a diagram illustrating an embodiment of a method of performing a residual transformation.

3 is a diagram illustrating an embodiment of a residual signal on which intra prediction is performed.

4 is a diagram illustrating an embodiment of a result of performing residual transformation.

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

DETAILED DESCRIPTION Hereinafter, exemplary embodiments of the present disclosure will be described in detail with reference to the accompanying drawings so that those skilled in the art may easily implement the present disclosure. As those skilled in the art would realize, the described embodiments may be modified in various different ways, all without departing from the spirit or scope of the present invention. In the drawings, parts irrelevant to the description are omitted for simplicity of explanation, and like reference numerals designate like parts throughout the specification.

Throughout this specification, when a portion is "connected" to another portion, this includes not only "directly connected" but also "electrically connected" with another element in between. do.

Throughout this specification, when a member is located “on” another member, this includes not only when one member is in contact with another member but also when another member exists between the two members.

Throughout this specification, when a part is said to "include" a certain component, it means that it can further include other components, without excluding the other components unless specifically stated otherwise. As used throughout this specification, the terms "about", "substantially" and the like are used at, or in the sense of, numerical values when a manufacturing and material tolerance inherent in the stated meanings is indicated, Accurate or absolute figures are used to assist in the prevention of unfair use by unscrupulous infringers. As used throughout this specification, the term "step to" or "step of" does not mean "step for."

Throughout this specification, the term “combination of these” included in the expression of the makushi form means one or more mixtures or combinations selected from the group consisting of the constituents described in the expression of the makushi form, wherein the constituents It means to include one or more selected from the group consisting of.

As an example of a method of encoding a real image and its depth information map, standardization is jointly performed by the Moving Picture Experts Group (MPEG) and the Video Coding Experts Group (VCEG), which have the highest coding efficiency among the video coding standards developed to date. Encoding may be performed by using high efficiency video coding (HEVC).

An embodiment of the present invention relates to a video encoding / decoding method and apparatus, and to a video encoding / decoding method and apparatus for improving coding performance by reducing a residual signal in encoding / decoding of block-based lossless intra prediction.

According to an embodiment of the present invention, residual transform and crossed residual transform are selectively performed in place of the existing intra prediction. The present invention can be applied only to the vertical or horizontal direction, and the existing prediction can be applied to other directions.

Here, the residual transform refers to performing the pixel unit DPCM in the prediction direction by using the residual signal that is the result of the intra prediction, and will be described with reference to FIG. 2.

FIG. 2 illustrates an embodiment of a method of performing pixel-by-pixel DPCM on a residual signal, and illustrates a vertical (mode 26) intra prediction of HEVC (High Efficiency Video Coding).

Referring to FIG. 2, l and q are used as reference pixels to predict the pixel p in the current block.

The residual signal is obtained by obtaining the difference between the pixel in the current block and the reference pixel according to the prediction direction. The equation for obtaining the residual signal of the first column is shown in Equation 1 below.

<Equation 1>

In Equation 1, r denotes a residual signal and may be represented as shown in FIG. 3.

Lossless intra prediction does not undergo transformation and quantization when the conventional prediction method is applied as described above, and thus the compression ratio decreases due to the size of the residual signal.

To solve this problem, a pixel unit DPCM using a pixel in the current block may also be used as a reference pixel, and may be expressed as in Equation 2 below.

<Equation 2>

In other words, the residual transform in the present invention is a method of maintaining a block-based processing structure during decoding while performing a pixel-by-pixel DPCM in a prediction direction using a residual signal that is a result of intra prediction, and is a block of HEVC (High Efficiency Video Coding). Unlike the base prediction, the residual signal is minimized by generating the residual signal by the difference between the nearest pixels in the prediction direction (intra prediction mode) by using the pixel inside the block to be currently predicted as the reference pixel.

On the other hand, the pixel-based DPCM may be in violation of the block-based processing because the pixel-based DPCM needs to be restored in the pixel unit. However, when restoring by using the residual transformation as shown in Equation 3 below, the pixel unit DPCM can be used while maintaining the block unit processing.

<Equation 3>

As a specific example, the encoder side applies the pixel unit DPCM in the same way as the intra prediction direction in the existing HEVC (application of the residual transform), and then calculates the residual signal as shown in Equation 3 to decode the residual signal on the decoder side ( Inverse residual transform allows for decoding while maintaining a block-based processing structure.

Since lossless intra compression does not undergo conversion and quantization unlike lossy intra compression, the residual signal occupies the largest portion of the coding efficiency. Therefore, neighboring pixels having a value most similar to the current pixel are used as reference pixels. It is advantageous for compression.

In addition, the present invention has an advantage that the block-based structure of the existing HEVC is not violated because the reconstruction is performed by using the residual signal r 'which is already known information transmitted from the encoder to the decoder without reconstruction delay which is a disadvantage of the pixel-by-pixel DPCM. .

In encoding, a residual signal obtained by applying a residual transform (pixel unit DPCM) as shown in Equation 2 is represented by r 'and may be represented as shown in FIG. 4.

The residual transformation is the horizontal direction (mode 10) in the conventional intra prediction. It can be applied in the vertical direction (mode 26) as described above, and in all other modes.

According to an embodiment of the present invention, the crossover residual transformation may be selectively performed after the residual transformation as described above.

In the present invention, the cross residual transform is obtained by applying the residual transform described above in the intra prediction direction to obtain the residual signal r 'and then applying the residual transform again in the direction intersecting the intra prediction direction with respect to the residual signal r'. To obtain the residual signal r ”.

After the above residual transformation process, performing the residual transformation as opposed to the prediction direction is a crossover residual transform, which can be expressed by Equation 4 below.

<Equation 4>

On the other hand, the above crossover residual transform can be selectively applied. For example, when the residual transformation is vertical, the crossover residual transformation is applied horizontally. If the residual transformation is horizontal, the crossover residual transformation may be applied vertically.

After performing the above cross residual transform, the frequency domain transform and the quantization process are omitted because they are a lossless compression method, and the entropy process may be the same as the existing intra video compression method (such as HEVC).

In the decoding process, the above process is reversed. In a block in which a residual prediction and a crossover residual transformation are performed after performing vertical prediction in the encoding process, a block-based inverse crossover residual transform is performed on r "transmitted to the decoder. (Inverse crossed residual transform) is applied and can be expressed by Equation 5 as follows.

<Equation 5>

R ', the result obtained after performing the inverse crossover residual transformation (horizontal direction) shown in Equation 5, can be decoded when the prediction reconstruction in the vertical direction is performed in the same manner as the above-described reconstruction of the block-based residual transformation.

The residual transform and the crossover residual transform of the present invention can be selectively applied. When the residual transform is applied during encoding, the inverse residual transform is applied during decoding. The intersection residual transformation will be applied.

According to an embodiment of the present invention, the selective execution of the residual transform and the crossover residual transform as described above is applied only to the vertical or horizontal direction, and the existing prediction method for other directions. This can be applied.

In the present invention, the selection of the residual transform and the cross residual transform is based on the rate-distortion optimization (RDO) (preferring two methods in the encoder, calculating the bit rate and the degree of distortion, and then selecting the (RD-Optimization) best performance). It can be divided into the case of one case and the case of using only one method through option setting in header.

In the case of the rate-distortion optimization (RDO) process, after applying both methods of the residual transformation and the crossover residual transformation, the method having the lower cost through the calculation of the rate-distortion optimization is selected.

In order to distinguish the selected method in the decoder, the encoder adds a signaling bit of 0 or 1 during mode transmission.

On the other hand, in the case of option setting in the header, the header additionally includes a bit to determine on / off regarding the use of the cross residual transform. If On is selected, the two methods are selected through the rate-distortion optimization described above. If Off is used, only residual transformation is used without additional signaling bits.

According to another embodiment of the present invention, residual transform and crossed residual transform may be selectively performed on the selected region instead of the existing intra prediction.

Herein, the selection area refers to designating a separate region such as an object or a block according to a specific purpose or characteristic of an image.

In addition, in order to decode that the residual transform or the cross residual transform is applied to the selection region, signaling bits are transmitted to the selection region during encoding.

In the case of signaling bit transmission for decoding, n bits may be used. In addition, various methods may be applied, such as a method of applying only a residual transform, a method of applying only a residual residual transform, or a method of selectively applying a residual transform and a cross residual transform to a selected region.

Meanwhile, the selection area is an area divided based on the existing object classification, extraction method, segmentation, ROI, etc., and may be divided in various ways in addition to the above-described method.

In the case of designating a selection area of a block, a random selection or a specific algorithm can be selected.

For example, designation of the selection area in the present invention may be divided into object designation and block designation.

First, in the case of the object designation, the residual transformation and the crossover residual transformation are applied only to a specific portion displayed by the object map in the state where the objects are distinguished through the object map or the like. It means to represent the image with 0 and 1.

That is, when an object is indicated by 1 in the object map, it is possible to apply the technique of the present invention without additional signaling bits.

In addition, when applying the technique of the present invention only to the inner region determined by the start position and end position, no additional signaling bit is required.

In the case of block designation, on / off signal bits for the technique of the present invention are transmitted to a block of a randomly selected LCU, CU, and PU.

For example, in case of 0, general encoding is applied, and in case of 1, residual transformation and cross residual transformation may be applied.

5 is a block diagram illustrating a configuration of an image encoding apparatus according to an embodiment of the present invention.

In general, the encoding apparatus includes an encoding process and a decoding process, and the decoding apparatus includes a decoding process. The decoding process of the decoding apparatus is the same as the decoding process of the encoding apparatus. Therefore, the following description focuses on the encoding apparatus.

As shown in FIG. 5, an image encoding apparatus according to an embodiment of the present invention includes coding units and structures, inter prediction, intra prediction, interpolation, filtering, and transformation. It includes a variety of new algorithms, including

Referring to FIG. 5, the apparatus for encoding an image may include an encoding mode determiner 110, an intra predictor 120, a motion compensator 130, a motion estimator 131, a transform encoder / quantizer 140, and an entropy encoding. The unit 150 includes an inverse quantization / conversion decoder 160, a deblocking filtering unit 170, a picture storage unit 180, a subtraction unit 190, and an adder 200.

The encoding mode determiner 110 analyzes the input video signal, divides the picture into coding blocks having a predetermined size, and determines an encoding mode for the divided coding blocks having a predetermined size. The encoding mode includes intra prediction encoding and inter prediction encoding.

The picture is composed of a plurality of slices, and the slice is composed of a plurality of largest coding units (LCUs). The LCU may be divided into a plurality of coding units (CUs), and the encoder may add information (flag) indicating whether to divide the bitstream. The decoder can recognize the location of the LCU using the address LcuAddr. The coding unit (CU) in the case where splitting is not allowed is regarded as a prediction unit (PU), and the decoder may recognize the location of the PU using a PU index.

The prediction unit PU may be divided into a plurality of partitions. In addition, the prediction unit PU may include a plurality of transform units (TUs).

The encoding mode determiner 110 transmits the image data to the subtractor 190 in a block unit (for example, PU unit or TU unit) of a predetermined size according to the determined encoding mode.

The transform encoding / quantization unit 140 converts the residual block calculated by the subtraction unit 190 from the spatial domain to the frequency domain. For example, two-dimensional discrete cosine transform (DCT) or discrete sine transform (DST) based transforms are performed on the residual block.

In addition, the transform encoding / quantization unit 140 determines a quantization step size for quantizing the transform coefficient, and quantizes the transform coefficient using the determined quantization step size. The quantization matrix may be determined according to the determined quantization step size and the encoding mode.

The quantized two-dimensional transform coefficients are transformed into one-dimensional quantized transform coefficients by one of a predetermined scanning method. The transformed sequence of one-dimensional quantized transform coefficients is supplied to the entropy encoder 150.

The inverse quantization / transform decoding unit 160 inverse quantizes the quantization coefficients quantized by the transform encoding / quantization unit 140. Furthermore, the inverse quantization coefficient obtained by inverse quantization is inversely transformed. Accordingly, the residual block transformed into the frequency domain may be restored to the residual block in the spatial domain.

The deblocking filtering unit 170 receives inverse quantized and inversely transformed image data from the inverse quantization / conversion encoder 160 and performs filtering to remove a blocking effect.

The picture storage unit 180 receives the filtered image data from the deblocking filtering unit 170 and restores and stores the image in picture units. The picture may be an image in a frame unit or an image in a field unit. The picture storage unit 180 includes a buffer (not shown) that can store a plurality of pictures. A number of pictures stored in the buffer are provided for intra prediction and motion estimation.

The pictures provided for intra prediction or motion estimation are called reference pictures.

The motion estimation unit 131 receives at least one reference picture stored in the picture storage unit 180 and performs motion estimation to output motion data including a motion vector, an index indicating a reference picture, and a block mode. do.

In order to optimize the prediction precision, the motion vector is determined with fractional pixel precision, for example 1/2 or 1/4 pixel precision. Since the motion vector may have fractional pixel precision, the motion compensation unit 130 applies an interpolation filter for calculating the pixel value of the fractional pixel position to the reference picture, whereby the pixel value of the fractional pixel position from the pixel value of the integer pixel position. To calculate.

The motion compensator 130 corresponds to a block to be encoded from a reference picture used for motion estimation among a plurality of reference pictures stored in the picture storage unit 180 according to the motion data input from the motion estimator 131. The prediction block is extracted and output.

The motion compensator 130 determines the filter characteristics of the adaptive interpolation filter required for the motion compensation with decimal precision. The filter characteristics are, for example, information indicating the filter type of the adaptive interpolation filter, information indicating the size of the adaptive interpolation filter, and the like.

The size of the filter is, for example, the number of taps that is the number of filter coefficients of the adaptive interpolation filter.

In detail, the motion compensator 130 may determine one of a split type and a non split type adaptive filter as the adaptive interpolation filter. The determined number of taps of the adaptive interpolation filter, and the value of each filter coefficient are then determined. The value of the filter coefficient may be determined differently for each position of the decimal pixel relative to the integer pixel. In addition, the motion compensation unit 130 may use a plurality of non-adaptive interpolation filters having a fixed filter coefficient.

The motion compensator 130 may set the characteristics of the interpolation filter in a predetermined processing unit. For example, it can be set in a decimal pixel unit, a coding basic unit (encoding unit), a slice unit, a picture unit, or a sequence unit. In addition, one characteristic may be set for one video data.

Therefore, since the same filter characteristic is used in the predetermined processing unit, the motion compensator 130 includes a memory for temporarily holding the filter characteristic. This memory retains filter characteristics, filter coefficients, and the like as necessary. For example, the motion compensator 130 may determine filter characteristics for each I picture and determine filter coefficients in units of slices.

The motion compensator 130 receives the reference picture from the picture storage unit 180 and applies a filter process using the determined adaptive interpolation filter to generate a predictive reference image with a small precision.

Then, the prediction block is generated by performing motion compensation with decimal pixel precision based on the generated reference image and the motion vector determined by the motion estimation unit 131.

The subtractor 190 receives a block in a reference picture corresponding to the input block from the motion compensator 130 and performs a difference operation with the input macroblock when the input block to be encoded is predictively encoded between the pictures. Output the (residue signal).

The intra predictor 120 performs intra prediction encoding by using the reconstructed pixel value inside the picture on which the prediction is performed. The intra prediction unit receives the current block to be predictively encoded and selects one of a plurality of preset intra prediction modes according to the size of the current block to perform intra prediction. The intra predictor 120 determines an intra prediction mode of the current block by using previously encoded pixels adjacent to the current block, and generates a prediction block corresponding to the determined mode.

The previously encoded region of the region included in the current picture is decoded again for use by the intra prediction unit 120 and stored in the picture storage unit 180. The intra predictor 120 generates a prediction block of the current block by using pixels adjacent to the current block or non-adjacent but applicable pixels in a previously encoded region of the current picture stored in the picture storage unit 180.

The intra predictor 120 may adaptively filter adjacent pixels to predict an intra block. For the same operation in the decoder, the encoder may transmit information indicating whether to filter. Alternatively, filtering may be determined based on the intra prediction mode of the current block and the size information of the current block.

The prediction type used by the image encoding apparatus depends on whether the input block is encoded in the intra mode or the inter mode by the encoding mode determiner.

The switching between the intra mode and the inter mode is controlled by the intra / inter switch.

The entropy encoder 150 entropy encodes the quantized coefficient quantized by the transform encoder / quantizer 140 and the motion information generated by the motion estimator 131. In addition, intra prediction mode, control data (eg, quantization step size, etc.) may be encoded. The filter coefficients determined by the motion compensator 130 are also encoded and output as a bit stream.

As described above, the configuration of the image decoding apparatus according to an embodiment of the present invention may be derived from the configuration of the image encoding apparatus illustrated in FIG. 5, for example, the inverse of the encoding process as described with reference to FIG. 5. By performing the process, the image can be decoded.

Meanwhile, in the lossless prediction according to an embodiment of the present invention, the transformation and quantization processes as described with reference to FIG. 5, and the filtering process such as a deblocking filter may be omitted.

The method according to the present invention described above may be stored in a computer-readable recording medium that is produced as a program for execution on a computer, and examples of the computer-readable recording medium include ROM, RAM, CD-ROM, magnetic tape , Floppy disks, optical data storage, and the like, and also include those implemented in the form of carrier waves (eg, transmission over the Internet).

The computer readable recording medium can be distributed over network coupled computer systems so that the computer readable code is stored and executed in a distributed fashion. In addition, functional programs, codes, and code segments for implementing the method can be easily inferred by programmers in the art to which the present invention belongs.

In addition, although the preferred embodiment of the present invention has been shown and described above, the present invention is not limited to the specific embodiments described above, but the technical field to which the invention belongs without departing from the spirit of the invention claimed in the claims. Of course, various modifications can be made by those skilled in the art, and these modifications should not be individually understood from the technical spirit or prospect of the present invention.

Claims (8)

In the method of performing intra prediction,
Generating a residual signal according to the intra prediction mode;
Performing residual transformation by applying a pixel unit DPCM in a first direction to the residual signal; And
Performing a cross residual transform by applying a pixel unit DPCM in a second direction to the residual transformed residual signal,
The residual transform and the cross residual transform are performed on a selected area based on a specific object or block. The method of claim 1, wherein the selected area is
An intra prediction method corresponding to at least one of a plurality of objects divided in an object map of an image. The method of claim 1,
The first direction is the same as a direction corresponding to the intra prediction mode, and the first direction and the second direction are directions perpendicular to each other. The method of claim 1,
And a direction corresponding to the intra prediction mode is a vertical or horizontal direction. The method of claim 1,
And performing the cross residual transform selectively. The method of claim 1,
And transmitting the information on the selected region to a decoding apparatus. In the apparatus for performing intra prediction,
A residual signal generator generating a residual signal according to the intra prediction mode;
A residual conversion performing unit performing residual conversion by applying a pixel unit DPCM in a first direction to the residual signal; And
A cross residual transform performing unit performing cross residual transformation by applying a pixel unit DPCM in a second direction to the residual transformed residual signal,
The residual transformation and the crossover residual transformation are performed on a region selected based on a specific object or block. The method of claim 7, wherein the selected area is
An intra prediction apparatus corresponding to at least one of a plurality of objects divided in an object map of an image.

Images