Analog VLSI Focal-Plane Array With Dynamic Connections for the Estimation of Piecewise-Smooth Optical Flow

Abstract

An analog very large-scale integrated (aVLSI) sensor is presented that is capable of estimating optical flow while detecting and preserving motion discontinuities. The sensor's architecture is composed of two recurrently connected networks. The units in the first network (the optical-flow network) collectively estimate two-dimensional optical flow, where the strength of their nearest-neighbor coupling determines the degree of motion integration. While the coupling strengths in our previous implementations were globally set and adjusted by the operator, they are now dynamically and locally controlled by a second on-chip network (the motion-discontinuity network). The coupling strengths are set such that visual motion integration is inhibited across image locations that are likely to represent motion boundaries. Results of a prototype sensor illustrate the potential of the approach and its functionality under real-world conditions.

On the implementation of linear diffusion in transconductance‐based cellular nonlinear networks

“…It is employed for image segmentation [1,2] and motion detection [3]. Beyond that, reaction-diffusion equations can be used to describe complex dynamics involving multiple state variables defined over a continuous or discretized plane or stack of planes [4].…”

Analog VLSI implementation of spatio-temporal frequency tuned visual motion algorithms

“…Performing this computation in special-purpose hardware is highly advantageous to reduce the load on later processing stages, potentially reducing power, weight, and size requirements over a conventional CCD/DSP combination. Very large-scale integration (VLSI) implementation of visual motion algorithms has been accomplished by a large number of authors over the last two decades [10]- [35] varying in implementation from digital block matching chips with serial inputs to analog VLSI motion sensors. One significant advantage of the analog VLSI approach for visual motion processing is continuous-time operation: the lack of temporal sampling removes the possibility of temporal aliasing.…”

Tracking improves performance of biological collision avoidance models

“…The size information can be accurately estimated only by distinguishing the object from the background which leads us to the very difficult problem of visual segmentation. Biological visual segmentation algorithms have been proposed (Koch et al 1986) and their electronic implementations have also been fabricated (Stocker 2004). However, these are complex solutions and require significant processing hardware and computational time.…”