Apparatus and method for performing rate-distortion optimization based on hadamard-quantization cost

Patent metadata

Field	Value
Publication number	US-2017195675-A1
Application number	US-201615015456-A
Country	US
Kind code	A1
Filing date	Feb 4, 2016
Priority date	Jan 5, 2016
Publication date	Jul 6, 2017
Grant date	—

Abstract

Disclosed herein are an apparatus and method for performing rate-distortion optimization based on cost. The encoding apparatus selects an encoding mode to be used to encode a target block from among multiple modes and performs computation for rate-distortion optimization in the encoding mode. The encoding apparatus calculates a cost of at least one of the multiple modes in relation to the encoding of the target block, and selects the encoding mode from among the multiple modes, based on the cost.

First claim

What is claimed is:

1 . An encoding apparatus, comprising:
a selection unit for selecting an encoding mode to be used to encode a target block from among multiple modes; and a computation unit for performing computation for rate-distortion optimization in the encoding mode, wherein the selection unit calculates a cost of at least one of the multiple modes in relation to encoding of the target block, and selects the encoding mode from among the multiple modes, based on the cost.

2 . The encoding apparatus of claim 1 , wherein the multiple modes comprise a skip mode, a merge mode, an inter mode, and an intra mode.

3 . The encoding apparatus of claim 1 , wherein:
the selection unit calculates costs of the multiple modes in relation to encoding of the target block, and the encoding mode is a mode having a minimum cost among the multiple costs.

4 . The encoding apparatus of claim 1 , wherein the cost is a similar rate-distortion cost that is a value related to rate-distortion cost when the mode is used for encoding the target block.

5 . The encoding apparatus of claim 1 , wherein the cost is calculated based on a model parameter for the corresponding mode, a Hadamard value in the mode, a Lagrangian multiplier in the mode, and a number of bits occurring in the mode.

6 . The encoding apparatus of claim 5 , wherein the Hadamard value is calculated based on both the target block and a prediction block for the target block.

7 . The encoding apparatus of claim 1 , wherein:
the cost is a sum of a first value and a second value, the first value is a product of a model parameter for the corresponding mode and a Hadamard value in the mode, and the second value is a product of a Lagrangian multiplier in the mode and an estimated value of a number of bits required for description of the mode.

8 . The encoding apparatus of claim 1 , wherein the cost of the mode is calculated without using a reconstructed block for the target block.

9 . The encoding apparatus of claim 1 , wherein:
the multiple modes comprise a merge mode and a skip mode, and the selection unit determines whether the encoding mode of the target block is the skip mode, using a prediction block in the merge mode and the target block.

10 . The encoding apparatus of claim 1 , wherein:
the multiple modes comprise a merge mode and a skip mode, and the selection unit determines the encoding mode of the target block to be the skip mode when a level of a Hadamard value in the merge mode is 0.

11 . The encoding apparatus of claim 1 , wherein the selection unit is configured to, when a level of a Hadamard value in the corresponding mode is equal to or greater than a predefined value, perform motion compensation at a depth subsequent to a current depth of the target block without performing motion compensation at the current depth.

12 . The encoding apparatus of claim 1 , wherein:
the multiple modes comprise an inter mode, and the selection unit determines whether to perform prediction in the inter mode upon encoding the target block, based on both a level of a Hadamard value in the inter mode and a first threshold function.

13 . The encoding apparatus of claim 1 , wherein:
the multiple modes comprise a skip mode, a merge mode, an inter mode, and an intra mode, the selection unit sets modes such that prediction in the skip mode and in the merge mode is performed, and such that prediction of the target block in the inter mode is performed while prediction of the target block in the intra mode is omitted if a level of a Hadamard value in the inter mode is in a high cost state, and the high cost state indicates a state in which the level of the Hadamard value in the inter mode has a value that causes prediction to be performed at a subsequent depth of the target block.

14 . The encoding apparatus of claim 13 , wherein the selection unit determines whether to perform prediction in the intra mode depending on a result of prediction in the inter mode.

15 . The encoding apparatus of claim 13 , wherein the selection unit determines whether to omit prediction in the intra mode and to perform prediction in the inter mode upon encoding the target block, based on both the level of the Hadamard value in the inter mode and a second threshold function.

16 . The encoding apparatus of claim 1 , wherein:
the multiple modes comprise a merge mode and an inter mode, and the selection unit determines whether to perform prediction in the inter mode, using both a prediction block in the merge mode and the target block.

17 . The encoding apparatus of claim 1 , wherein:
the multiple modes comprise a merge mode and an inter mode, and the selection unit determines whether to perform prediction in the inter mode, using both a prediction block in the merge mode and the target block if a Hadamard value in the merge mode is in a high cost state.

18 . The encoding apparatus of claim 1 , wherein:
the multiple modes comprise a skip mode, a merge mode, an inter mode, and an intra mode, and the selection unit determines a mode in which motion compensation is to be performed, among the multiple modes, using both the target block and a prediction block for the target block if levels of Hadamard values in the multiple modes are identical to each other.

19 . The encoding apparatus of claim 1 , wherein the computation unit does not perform computation for rate-distortion optimization in modes other than a mode selected as the encoding mode, from among the multiple modes.

20 . An encoding method, comprising:
selecting an encoding mode to be used to encode a target block from among multiple modes; and performing computation for rate-distortion optimization in the encoding mode, wherein selecting the encoding mode comprises: calculating a cost of at least one of the multiple modes in relation to encoding of the target block; and selecting the encoding mode from among the multiple modes, based on the cost.

Assignees

• Electronics & Telecommunications Res Inst

Inventors

• Seok Jin-Wuk
• Ki Myung-Seok
• Kim Youn-Hee
• Kim Hui-Yong
• Choi Jin-Soo

Classifications

• H04N19/103

  Selection of coding mode or of prediction mode · CPC title

• H04N19/147 Primary

  according to rate distortion criteria (rate-distortion as a criterion for motion estimation H04N19/567) · CPC title

• H04N19/15

  by monitoring actual compressed data size at the memory before deciding storage at the transmission buffer · CPC title

• H04N19/176

  the region being a block, e.g. a macroblock · CPC title

• H04N19/61

  in combination with predictive coding · CPC title