A Process for Sensor Configuration in Multi-Camera Networks

Multi-camera networks are an increasingly common approach to three dimensional sensing. Although the potential of this technology is appealing, designing such networks is a difficult and still open problem. How many and what type of cameras should be used? How should they be distributed in the environment? In this report we explore how these design parameters influence system performance using Monte Carlo techniques. We also consider the problem of optimal sensor placement by investigating a novel approach to optimization using genetic algorithms, and demonstrate the results through simulation.