Graph-based visual sensor deployment and optics optimization

In this brief, we present a method for automatic deployment of visual sensor networks. The solution is formulated as a greedy search by taking advantage of the topological properties of the vision graph. The optimization of the optical component of the vision system is solved using a geometrical approach. We show that the information about the resolution, focus, and visibility of the vision system can be encoded in the geometry of a square frustum. The frustum is used to model the optimal configuration of the lens. Finally, we make use of experiments to show the effectiveness of our approach.