Part of Advances in Neural Information Processing Systems 3 (NIPS 1990)
Wyeth Bair, Christof Koch
We have designed and tested a one-dimensional 64 pixel, analog CMOS VLSI chip which localizes intensity edges in real-time. This device exploits on-chip photoreceptors and the natural filtering properties of resistive net(cid:173) works to implement a scheme similar to and motivated by the Difference of Gaussians (DOG) operator proposed by Marr and Hildreth (1980). Our chip computes the zero-crossings associated with the difference of two ex(cid:173) ponential weighting functions. If the derivative across this zero-crossing is above a threshold, an edge is reported. Simulations indicate that this technique will extend well to two dimensions.