An image sensor with fast objects' positions extraction function

Abstract

This paper describes a kind of vision chip, which is an integration of an image processing circuit with photo receptors, that has a function of extracting objects' positions in focal plane. The objects' positions are output as their coordinates, which are useful for further detailed image recognition processing. The extraction processing has two steps; first, the flags indicating the objects' center positions are generated by analog parallel processing circuit implemented by resistive network and comparators, and next, the coordinates of such flags are generated by-and-priority encoders and a novel successive masking circuit. The designed circuit is capable of centroid detection for 23 23 pixels within 50 s, and the processing time is expected not to increase so much even if the number of pixels increases, which will represent an improvement over the conventional processor-based image processing system. We also proposed and evaluated another implementation of a centroid detector using pulse-width adder.

High speed vision processor with reconfigurable processing element array based on full-custom distributed memory

“…Original vision chips integrate a massively parallel processing architecture and an image sensor to achieve high speed early visual functions as the vertebrate retina does. [3][4][5][6][7][8][9][10][11][12][13][14][15] Such applications include object position extraction, [3][4][5][6][7] image filtering, [8][9][10][11][12] contour generation, 13) dynamic range extension, 14) and event detection. 15) More advanced vision chip designs focus on more specific visual tasks such as fingerprint recognition, 16) eye tracking 17) and horizon detection.…”

A micropower vision processor for parallel object positioning and sizing

“…Other systems have combined this with additional functionality, for example, to combine centroiding with motion detection [3] [4] or embed an APS imager [5]. Multiple centroid capability has been achieved by using a window and search type algorithm to direct the COM computation [6] [7]. This paper reports the first vision chip capable of objectposition detecting and sizing of unlimited objects in genuine parallelism.…”