Method and apparatus using software engine and hardware engine collaborated with each other to achieve hybrid video encoding

• US 10,057,590 B2
• Filed: 09/15/2016
• Issued: 08/21/2018
• Est. Priority Date: 01/13/2014
• Status: Active Grant

• Current Assignee: Xueshan Technologies Inc. (Wi-LAN Inc.)
• Original Assignee: MediaTek, Inc.
• Inventors: Huang, Chao-Chih, Lin, Ting-An, Chang, Shen-Kai, Chou, Han-Liang
• Primary Examiner(s): Li, Tracy Y
• Application Number:
  US15/265,896
• Publication Number:
  US 20170006294A1
• Time in Patent Office: 705 Days
• CPC Class Codes: H04N 19/42 characterised by implementation details or hardware specially adapted for video compression or decompression, e.g. dedicated software implementation H04N19/635 takes precedence
• CPC Class Codes: H04N 19/423 characterised by memory arrangements H04N19/433 takes precedence
• CPC Class Codes: H04N 19/43 Hardware specially adapted for motion estimation or compensation
• CPC Class Codes: H04N 19/433 characterised by techniques for memory access
• CPC Class Codes: H04N 19/439 using cascaded computational arrangements for performing a single operation, e.g. filtering

Abstract

A hybrid video encoding method and system using a software engine and a hardware engine. The software engine receives coding unit data associated with a current picture, and performs a first part of the video encoding operation by executing instructions. The first part of the video encoding operation generates an inter predictor and control information corresponding to the coding unit data of the current picture. The first part of the video encoding operation stores the inter predictor into an off-chip memory. The hardware engine performs a second part of the video encoding operation according to the control information. The second part of the video encoding operation receives the inter predictor, and subtracts the inter predictor from the coding unit data to generate a residual signal. The second part of the video encoding operation then transforms and quantizes the residual signal to generate transformed and quantized residual signal, and encodes the transformed and quantized residual signal to generate an encoded video bitstream.

Claims

1. A method of encoding video data in a video coding system, comprising:

receiving coding unit data associated with a current picture;

performing a first part of a video encoding operation by a software engine to execute a plurality of instructions, wherein the first part of the video encoding operation comprises;

generating an inter predictor and control information corresponding to the coding unit data of the current picture, wherein the inter predictor comprises image blocks for luma component and chroma components; and

storing the inter predictor into a memory buffer; and

performing a second part of the video encoding operation according to the control information by a hardware engine, wherein the second part of the video encoding operation comprises;

receiving the inter predictor from the memory buffer;

subtracting the inter predictor from the coding unit data to generate a residual signal;

transforming, quantizing the residual signal to generate transformed and quantized residual signal; and

encoding the transformed and quantized residual signal to generate an encoded video bitstream.

2. The method of claim 1, wherein the step of performing the first part of the video encoding operation comprises:

determining a description of the coding unit data in the current picture, wherein Me description comprises mode information, motion information, or both mode and motion information for the coding unit; and

including the description in the control information and storing the control information in a cache buffer.

3. The method of claim 1, wherein the step of performing the first part of the video encoding operation comprises:

determining at least one adjustable header parameter for the current picture, wherein the adjustable header parameter comprises filtering information, reference picture indication, or both filter information and reference picture indication; and

including the adjustable header parameter in the control information and storing the control information in a cache buffer.

4. The method of claim 1, wherein the second part of the video encoding operation further comprises one or a combination of an intra prediction operation, an inverse transform operation, an inverse quantization operation, a post processing operation, and a loop filter operation;

wherein the hardware engine further generates one or more of reconstructed data corresponding to the current picture, and side information or statistics of the current picture.

5. The method of claim 1, wherein the first part of the video encoding operation comprises performing motion estimation and motion compensation on source frame data of the current picture using a reconstructed picture as a reference picture.

6. The method of claim 1, wherein the control information comprises mode information, motion information, quantization information, block partition structure, in-loop filter parameters, or a combination thereof.

7. The method of claim 1, wherein at least a portion of the control information generated by the software engine is incorporated in the encoded video bitstream.

8. A hybrid video encoder, comprising

a software engine, arranged for performing a first part of a video encoding operation by executing a plurality of instructions, wherein the first part of the video encoding operation comprises:

generating an inter predictor and control information corresponding to a coding unit data of the current picture, wherein the inter predictor comprises image blocks for luma component and chroma components; and

storing the inter predictor into a memory buffer; and

a hardware engine, coupled to the software engine, the hardware engine arranged for performing a second part of the video encoding operation according to the control information, wherein the hardware engine comprises;

an adder, coupled to the memory buffer, for receiving the inter predictor and subtracting the inter predictor from the coding unit data to generate a residual signal;

a transformation circuit, couple to the adder, for transforming the residual signal to generate a transformed residual signal;

a quantization circuit, coupled to the transformation circuit, for quantizing the transformed residual signal to generate transformed and quantized residual signal; and

an entropy encoder, couple to the quantization circuit, for encoding the transformed and quantized residual signal to generate an encoded video bitstream.

9. The hybrid video encoder of claim 8, wherein the software engine determines a description of the coding unit data in the current picture as the control information, the description comprises mode information, motion information, or both the mode and motion information for the coding unit.

10. The hybrid video encoder of claim 8, wherein the software engine determines at least one adjustable header parameter for the current picture as the control information, the adjustable header parameter comprises filtering information, reference picture indication, or both the filter information and reference picture indication.

11. The hybrid video encoder of claim 8, wherein the second part of the video encoding operation further comprises one or a combination of an intra prediction circuit, an inverse transform circuit, an inverse quantization circuit, a post processing circuit, and a loop filter circuit;

wherein the hardware engine further generates one or more of reconstructed data corresponding to the current picture, and side information or statistics of the current picture.

12. The hybrid video encoder of claim 8, wherein the first part of the video encoding operation comprises performing motion estimation and motion compensation on source frame data of the current picture using a reconstructed picture as a reference picture.

13. The hybrid video encoder of claim 8, wherein the control information comprises mode information, motion information, quantization information, block partition structure, in-loop filter parameters, or a combination thereof.

14. The hybrid video encoder of claim 8, wherein at least a portion of the control information generated by the software engine is incorporated in an encoded video bitstream.