A CMOS imager with on-chip processing for image enhancement and edge detection
Faculty of Computing, Health and Science
School of Engineering and Mathematics
In this paper a CMOS vision chip featuring real-time on-chip image processing capabilities is presented. A sensor array, an analogue read-out processor and a digital image processor have been integrated on a single chip. The prototype can perform dynamic-range compression, noise filtering, edge detection and image enhancement. To achieve high-speed real-time processing, an efficient VLSI implementation based on a mixed-mode analogue-digital strategy is proposed. The proposed approach also leads to a very compact VLSI implementation of the analogue processing parts, allowing for 75% of the silicon area to be fully dedicated to the 58 × 58-pixel array.