DocumentCode :
1712383
Title :
Instruction set extension for high throughput disparity estimation in stereo image processing
Author :
Banz, Christian ; Dolar, Carsten ; Cholewa, Fabian ; Blume, Holger
Author_Institution :
Inst. of Microelectron. Syst., Leibniz Univ. Hannover, Hannover, Germany
fYear :
2011
Firstpage :
169
Lastpage :
175
Abstract :
This paper presents the implementation and evaluation of an application-specific instruction set for a customizable RISC-processor for very high throughput stereo image processing. Compared to the base processor the overall processing time is accelerated by a factor of over 130, while the processor silicon area requirement increases only by a factor of 2.9. The processor has been enhanced with algorithm-specific extensions (i.e. special functional units), as well as with extensions that are not restricted to a specific algorithm (e.g. single-instruction-multiple-data). Hereby, the special functional units account for 50% of the speed-up, but less than 14% of the processor silicon area requirement. The proposed processor extensions thereby sustain the full flexibility of a programmable processor while enabling disparity estimation of 640×480 stereo video sequences at 20 fps when running at a clock frequency of 373 MHz.
Keywords :
application specific integrated circuits; image sequences; reduced instruction set computing; stereo image processing; application specific instruction set; clock frequency; customizable RISC processor; high throughput disparity estimation; instruction set extension; processor silicon area requirement; programmable processor; stereo image processing; stereo video sequences; Acceleration; Algorithm design and analysis; Estimation; Hardware; Registers; Signal processing algorithms; Transforms; ASIP; custom instructions; semi-global matching; stereo video processing;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Application-Specific Systems, Architectures and Processors (ASAP), 2011 IEEE International Conference on
Conference_Location :
Santa Monica, CA
ISSN :
2160-0511
Print_ISBN :
978-1-4577-1291-3
Electronic_ISBN :
2160-0511
Type :
conf
DOI :
10.1109/ASAP.2011.6043265
Filename :
6043265
Link To Document :
بازگشت