System-on-a-chip design of a low-power smart vision system

A low power smart imager design is proposed for real time machine vision applications. It takes advantages of recent advances in integrated sensing/processing designs, electronic neural networks, and sub-micron VLSI technology. The smart vision system integrates an active pixel camera with a programmable neural computer and an advanced microcomputer. A system-on-a-chip implementation of this smart vision system is shown to be feasible by integrating the whole system into a 3-cm×3-cm chip design in a 0.18 m CMOS technology. The on-chip neural computer provides one tera-operation-per-second computing power for various parallel vision operations and smart sensor functions. Its high performance is due to massively parallel computing structures, high data throughput rates, fast learning capabilities, and system-on-a-chip implementation. This highly integrated smart imager can be used for various scientific missions and other military, industrial or commercial vision applications.