Title :
An edge enhancement technique for analog VLSI early vision applications
Author :
Rowley, Matthew D. ; Harris, John G.
Author_Institution :
Elgin AFB, USA
Abstract :
Several efforts have been made previously by researchers attempting to mimic biological processing in mammalian retinas to perform such tasks as motion detection/analysis, image requisition, and object position and center of gravity determination. In all these efforts, performance was limited by offsets in the analog circuitry. This paper will discuss a technique referred to as nonnearest neighbour local differencing which improves image feature detection capabilities in low contrast images or noisy environments
Keywords :
VLSI; analogue integrated circuits; digital signal processing chips; edge detection; feature extraction; image sensors; analog VLSI early vision applications; center of gravity determination; edge enhancement technique; image feature detection; image requisition; low contrast images; mammalian retinas; motion analysis; motion detection; noisy environments; nonnearest neighbour local differencing; object position determination; Circuits; Computer vision; Gravity; Image analysis; Image edge detection; Image motion analysis; Motion detection; Performance analysis; Retina; Very large scale integration;
Conference_Titel :
Neural Networks, 1996., IEEE International Conference on
Conference_Location :
Washington, DC
Print_ISBN :
0-7803-3210-5
DOI :
10.1109/ICNN.1996.549012
