Solving Surveillance Coverage Demand Based on Dynamic Programming

Abstract

Typical visual sensor planning approaches install visual sensors arrays to increase the amount of coverage and/or decrease the installation cost. These planning approaches operate with no stress on coverage demand, thus, optimizing the visual sensor placement based on equally significance grids. This paper addresses the visual sensor coverage efficiency based on a combination of risk mapping and dynamic programming. The improved coverage efficiency is obtained by utilizing a prior routine to highlight the security critical regions. Then, a dynamic programming algorithm is used to compute a near optimal coverage solution. The result of the dynamic programming is evaluated with respect to global greedy search outcomes. The comparison reveals the reliability of the visual sensor planning using risk maps and dynamic programming.