KR20120107386A - Method for determining fast mode using prediction motion vector of scalable video coding extended inter layer - Google Patents

Abstract

PURPOSE: A method for determining a high-speed mode using a prediction motion vector between scalable video coding extended inter layers is provided to store a coding value of a part around a basic layer, thereby determining a prediction vector in consideration of correlation between a current basic layer and a surrounding macro block. CONSTITUTION: A high-speed mode determining device operates a motion vector by coding a current basic layer block(S201). The device operates a motion vector by coding a basic layer block adjacent to the current basic layer block(S202). The device determines correlation based on the current basic layer block and the motion vector of a basic layer block adjacent to the current basic layer block(S203). The device determines a motion vector prediction omission mode of the current basic layer block according to the determination result(S204). [Reference numerals] (S201) Coding a basic layer; (S202) Coding the basic layer block of a peripheral block; (S203) Determining the correlation of the basic layer and the basic layer of the peripheral block; (S204) Basic layer omitting mode; (S205) Inter prediction(16x8,8x16,8x8); (S206) RD cost(16x8,8x16,8x8); (S207) Mode prediction(16x16); (S208) RD cost(16x16); (S209) RD optimum mode decision

Description

Fast Mode Determination using Scalable Motion Coding Spatial Inter-layer Predictive Motion Vectors {METHOD FOR DETERMINING FAST MODE USING PREDICTION MOTION VECTOR OF SCALABLE VIDEO CODING EXTENDED INTER LAYER}

The present invention relates to a fast mode determination method using a predictive motion vector between scalable image coding spatial hierarchies. More particularly, the present invention relates to a method for determining a prediction vector in consideration of correlations between neighboring macroblocks as well as current macroblocks. The present invention also relates to a fast mode determination method using a scalable video encoding spatial inter-layer prediction motion vector capable of improving image quality.

In general, scalable video coding (SVC) is a video codec capable of providing a hierarchically varying bitstream in terms of temporal, spatial and image quality according to network conditions or terminal performance in various multimedia environments. to be.

Since the change in temporal, spatial and image quality can be made by removing or selecting only some bitstreams from one bitstream, the transmission and consumption steps are very effective in various multimedia environments.

In other words, a decoded bitstream of arbitrary points on three axes of time, space, and image quality is extracted and decoded from one coded bitstream having the image quality, the highest frame rate, and the maximum image size. , An image having a frame rate can be obtained.

H.264 / SVC is an extension standard of H.264 / AVC. It was developed and completed by JVT (Joint Video Team) jointly formed by ISO / IEC MPEG and ITU-T's VCEG.

Previously, scalability for video coding has been attempted in MPEG-2, H.263, and MPEG-4 visuals. However, the scalability of the video coding is not used because it has low coding efficiency and does not adapt adaptively in various environments. It was.

H.264 / SVC can be combined with various scalability features compared to the existing scalable video codec, so it is highly adaptable to the environment, and various inter-layer reference modes are added, so that it is competitive in performance with a single layer encoder. Video codec.

Scalable video coding is an extension of the H.264 / AVC standard, supporting temporal, spatial and image quality scalability.

Among them, spatial scalability is a method of predicting using an interlayer motion vector.

In order to perform spatial scalability, prediction is performed by using information of a lower layer, and encoding is performed based on this information.

At this time, since encoding and selecting information for minimizing the bit rate distortion value RD-Cost are encoded for each piece of information, encoding efficiency can be increased, but a large amount of computation is required due to repetitive operations.

The technical structure described above is a background technique for assisting the understanding of the present invention, and does not mean the prior art widely known in the technical field to which the present invention belongs.

In the prior art, in order to determine a fast mode in the spatial enhancement layer, the macroblock considers that the motion vector has a value of zero for most of the test images, so that the enhancement layer diagram of the same spatial point when the motion vector is zero Considering the 16x16 block, the enhancement layer reduces the complexity in the enhancement layer by obtaining the motion vector prediction and the bit rate distortion only in the 16x16 block mode.

However, such a prior art has to consider the motion vector obtained from the base layer and the vector corrected to the 1/4 pixel precision level, and for the various images, the current image of the zero motion block is considered in the base layer. The motion vector is not accurate, which causes a problem of degrading the image quality.

The present invention provides a fast mode decision method using a predictive motion vector between scalable image coding spatial hierarchies that can improve image quality for various images by determining a prediction vector in consideration of correlations between not only current macro blocks but also neighboring macro blocks. The purpose is to provide.

According to the present invention, a method of determining a fast mode using a scalable video encoding spatial inter-layer prediction motion vector includes: calculating a motion vector by coding a current base layer block; Calculating a motion vector by coding a base layer block adjacent to the current base layer block; And determining an association based on a motion vector of the current base layer block and the neighboring base layer block, and determining a motion vector prediction skip mode of the current base layer block according to the determination result. It is done.

In the present invention, the determining of the motion vector prediction skipping mode of the current base layer block includes: calculating a median value of a motion vector of a neighboring base layer block adjacent to the current base layer block; And if the calculated intermediate value is a zero vector, entering the motion vector prediction skipping mode of the current base layer block.

In the present invention, if the intermediate value is a zero vector, entering a motion vector prediction skipping mode and performing inter prediction on a mode of 16 × 16 to calculate a bit rate distortion value (RD Cost); And determining a bit rate distortion optimal mode according to the calculated bit rate distortion value.

In the present invention, when the intermediate value is not a zero vector, performing the inter prediction for the modes of 16x8, 8x16, 8x8 to calculate the bit rate distortion value; And determining a bit rate distortion optimal mode according to the calculated bit rate distortion value.

In addition, the fast mode determination apparatus using the scalable video encoding spatial inter-layer prediction motion vector according to the present invention comprises: a base layer coding unit for coding a current base layer block; A peripheral block coding unit coding a base layer block adjacent to the current base layer block; An association determination unit that determines an association between the current base layer block and the neighboring base layer block; And a motion vector prediction omit mode determiner configured to determine a motion vector prediction omit mode according to the determination result of the association determining unit.

In the present invention, if the motion vector prediction skip mode determination unit determines the motion vector prediction skip mode, the mode predictor for performing inter prediction for the 16x16 mode; And an RD cost unit for calculating a bit rate distortion value (RD Cost) for the 16x16 mode.

In the present invention, when the motion vector prediction skip mode determination unit determines the normal motion vector prediction mode, the inter prediction unit for performing inter prediction on the 16x8, 8x16, 8x8 mode; And an RD cost unit for calculating bit rate distortion values for the 16x8, 8x16, and 8x8 modes, respectively.

The method may further include an RD optimum mode determiner configured to determine a bit rate distortion optimal mode according to the calculated bit rate distortion value.

As described above, the present invention saves the value of the peripheral coding for the base layer to determine the prediction vector in consideration of the correlation between the current base layer and the neighboring macroblock, thereby reducing the error of prediction and performing in various test images. And an effect of improving image quality.

FIG. 1 is a block diagram illustrating a configuration of a fast mode determination apparatus using a scalable video encoding spatial inter-layer prediction motion vector according to an embodiment of the present invention.
2 is a flowchart illustrating a fast mode determination method using a scalable video encoding spatial inter-layer prediction motion vector according to an embodiment of the present invention.
3 is a view for explaining a prediction method for a base layer according to an embodiment of the present invention.

Hereinafter, the present invention will be described in more detail with reference to Examples. These embodiments are only for illustrating the present invention, and the scope of rights of the present invention is not limited by these embodiments.

In this process, the thicknesses of the lines and the sizes of the components shown in the drawings may be exaggerated for clarity and convenience of explanation. In addition, the terms described below are defined in consideration of the functions of the present invention, which may vary depending on the intention or custom of the user, the operator. Therefore, definitions of these terms should be made based on the contents throughout this specification.

1 is a block diagram illustrating a configuration of a fast mode determination apparatus using a scalable video encoding spatial inter-layer prediction motion vector according to an embodiment of the present invention, and FIG. 2 is scalable according to an embodiment of the present invention. FIG. 3 is a flowchart for describing a fast mode determination method using a predictive motion vector between video encoding spatial layers. FIG. 3 is a diagram for describing a prediction method for a base layer according to an embodiment of the present invention.

The present invention provides a method of calculating a motion vector using median values of neighboring blocks in order to estimate a motion vector of a base layer block in a prediction technique with a large amount of scalable video coding (SVC) calculation.

1 to 3, a fast mode determination apparatus using a scalable video encoding spatial inter-layer prediction motion vector according to an embodiment of the present invention includes a base layer coding unit 1 for coding a current base layer block. Peripheral block coding section 2 for coding the base layer block adjacent to the current base layer block.

In addition, the association determination unit 3 that determines the association between the current base layer block and the neighboring base layer block, and the motion vector prediction skip mode that determines the motion vector prediction skip mode according to the determination result of the association determination unit 3 are determined. Part 4 is included.

The association determination unit 3 determines an association based on a current vector of the base layer block and the motion vector of the neighboring base layer block, and the prediction skip mode determination unit 4 determines the current association according to the determination result of the association determination unit 3. A motion vector prediction skip mode of the base layer block is determined.

In addition, when the motion vector prediction skip mode determination unit 4 determines the motion vector prediction skip mode, the mode predictor 5 performs inter prediction on the 16x16 mode, and the bit rate distortion value RD Cost for the 16x16 mode. RD cost part 7 which computes ()) is included.

In addition, when the motion vector prediction skipping mode determination unit 4 determines the normal motion vector prediction mode, the inter prediction unit 6 which performs inter prediction on the modes of 16x8, 8x16, and 8x8, and 16x8, 8x16, and 8x8 An RD cost section 8 for calculating bit rate distortion values for each mode is included.

Also included is an RD optimum mode determiner 9 for determining the bit rate distortion optimal mode according to the calculated bit rate distortion value.

A fast mode determination method using a scalable video encoding spatial inter-layer prediction motion vector according to an embodiment of the present invention configured as described above is as follows.

First, a motion vector is calculated by coding a current base layer block (S201), and at the same time, a motion vector is calculated by coding a base layer block adjacent to the current base layer block.

The motion vector operation of the neighboring neighboring base layer blocks stores the motion vectors of the current neighboring block and neighboring blocks in a register as shown in FIG. 3. Thereafter, the association is determined based on the motion vector of the current base layer block and the neighboring base layer block (S203).

The correlation determination determines the correlation by calculating the median value based on the motion vector of the current basic layer block and the motion vector of the neighboring block, and determines the skip mode of motion vector prediction of the current basic layer block according to the result of the correlation determination. S204).

The motion vector prediction skip mode of the current base layer block is determined by entering a motion vector prediction skip mode of the base layer block if the median value of the motion vector of the neighboring base layer block is zero. do.

If the intermediate value is not a zero vector, inter prediction is performed for modes of 16x8, 8x16, and 8x8 (S205), and a bit rate distortion value RD-Cost is calculated (S206).

If the median value is a zero vector, the motion vector prediction skip mode is entered and inter prediction is performed in a mode of 16x16 (S207) to calculate a bit rate distortion value (RD Cost) (S208).

Next, the bit rate distortion optimal mode is determined according to the calculated bit rate distortion value (S209).

In other words, if the intermediate value of the motion vector of the neighboring basic layer block is zero, the inter prediction is performed only for 16x16. Otherwise, the interlayer prediction is performed in 16x8, 8x16, and 8x8 modes. An inter prediction is performed for the bit rate distortion value. Finally, the RD optimal mode is determined based on each bit rate distortion value.

As such, the present invention has the effect of improving the accuracy of the reconstructed image by increasing the accuracy of the predicted value, which is an existing problem, by adopting the median of the neighboring blocks to increase the accuracy of the prediction. This has the effect of improving the image quality by 0.3dB compared to the existing method.

In summary, the present invention can reduce the error of prediction by storing the value of the peripheral coding for the base layer to determine the prediction vector in consideration of the correlation between the current base layer and the neighboring macroblock, and the performance and image quality in various test images. Can be improved.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is to be understood that the invention is not limited to the disclosed embodiments, but, on the contrary, It is understandable. Accordingly, the true scope of the present invention should be determined by the following claims.

1: base layer coding unit 2: peripheral block coding unit
3: Association determining unit 4: Motion vector skip mode determination unit
5: mode prediction unit 6: inter prediction unit
7,8: RD cost part 9: RD optimum mode decision part

Claims (1)

Calculating a motion vector by coding a current base layer block;
Calculating a motion vector by coding a base layer block adjacent to the current base layer block; And
Determining an association based on a motion vector of the current base layer block and the neighboring base layer block, and determining a motion vector prediction skip mode of the current base layer block according to a determination result. A fast mode determination method using scalable video coding spatial inter-layer prediction motion vectors.

Images