Method and system for spatial prediction in a video encoder
View Patent ↗Described herein is a method and system for spatial prediction in a video encoder. The method and system can operate on a set of video elements in parallel. Multiple modes can also be executed for each set of video elements. Modes correspond to a weighted combination of neighboring video elements. The weighted combination can be changed and loaded into a spatial predictor instruction table.
1. A method for spatial prediction in a video encoder, said method comprising:
storing a plurality of instructions in a table, wherein an instruction performs a correlation between one or more predetermined weights and one or more video samples; and
executing a current instruction with a controller, thereby producing a spatial estimate of a current video sample in a current mode.
2. The method of claim 1 , wherein the table contains instructions for 9 spatial estimation modes.
3. The method of claim 1 , wherein the current mode is evaluated based on a sum of absolute difference.
4. The method of claim 1 , wherein the table contains instructions for 16 video sample positions.
5. The method of claim 1 , wherein the current instruction is selected according to the current mode and a relative position of the current video sample.
6. The method of claim 1 , wherein more than one instruction is selected at one time.
7. The method of claim 5 , wherein 16 instructions are selected at one time.
8. A system for spatial prediction in a video encoder, said system comprising:
an instruction table storing a plurality of instructions, wherein an instruction performs a correlation between one or more predetermined weights and one or more video samples; and
an engine for executing a current instruction, thereby producing a spatial estimate of a current video sample in a current mode.
9. The system of claim 8 , wherein the table contains instructions for 9 spatial estimation modes.
10. The system of claim 8 , wherein the system further comprises:
a mode decision engine for evaluating the current mode based on a sum of absolute difference.
11. The system of claim 8 , wherein the table contains instructions for 16 video sample positions.
12. The system of claim 8 , wherein the engine further selects the current instruction according to the current mode and a relative position of the current video sample.
13. The system of claim 8 , wherein the engine executes more than one instruction at one time.
14. A system for spatial prediction in a video encoder, said system comprising:
an integrated circuit comprising:
a memory storing a plurality of instructions, wherein an instruction performs a correlation between one or more predetermined weights and one or more video samples; and
a first circuit for executing a current instruction, thereby producing a spatial estimate of a current video sample in a current mode.
15. The system of claim 14 , wherein the memory contains instructions for 9 spatial estimation modes.
16. The system of claim 14 , wherein the system further comprises:
a second circuit for evaluating the current mode based on a sum of absolute difference.
17. The system of claim 14 , wherein the memory contains instructions for 16 video sample positions.
18. The system of claim 14 , wherein the system further comprises:
a third circuit for selecting the current instruction according to the current mode and a relative position of the current video sample.
19. The system of claim 14 , wherein the first circuit executes more than one instruction at one time.
20. The system of claim 14 , wherein the first circuit executes 16 instructions in parallel.