Title :
Customizing CPU Instructions for Embedded Vision Systems
Author :
Piskorski, Stephane ; Lacassagne, Lionel ; Bouaziz, Samir ; Etiemble, Daniel
Author_Institution :
Univ. Paris Sud, Paris
Abstract :
This paper presents the customization of two processors: the Altera NIOS2 and the Tensilica Xtensa, for fundamental algorithms in embedded vision systems: the salient point extraction and the optical flow computation. Both can be used for image stabilization, for drones and autonomous robots. Using 16-bit floating-point instructions, the architecture optimization is done in terms of accuracy, speed and power consumption. A comparison with a PowerPC Altivec is also done.
Keywords :
computer vision; digital signal processing chips; image sequences; 16-bit floating-point instructions; Altera NIOS2; PowerPC Altivec; Tensilica Xtensa; autonomous robots; drones; embedded vision systems; image stabilization; optical flow computation; salient point extraction; Computer architecture; Computer vision; Embedded system; Energy consumption; Image motion analysis; Iterative algorithms; Machine vision; Optical computing; Optical signal processing; Signal processing algorithms;
Conference_Titel :
Computer Architecture for Machine Perception and Sensing, 2006. CAMP 2006. International Workshop on
Conference_Location :
Montreal, Que.
Print_ISBN :
978-1-4244-0685-2
Electronic_ISBN :
978-1-4244-0686-9
DOI :
10.1109/CAMP.2007.4350352
