Optimizing Visual Sensors Placement With Risk Maps Using Dynamic Programming

Typically, optimizing the poses and placement of surveillance cameras is usually formulated as a discrete combinatorial optimization problem. The traditional aspects of solving the camera placement problem attempt to maximize the area monitored by the camera array and/or reduce the cost of installing a set of surveillance cameras. Several approximate optimization techniques have been proposed to locate near-optimal solution to the placement problem. Thus, related surveillance planning methods optimize the placement of visual sensors based on equally significance grids by not limiting to demand of coverage. This article explores the efficiency of the visual sensor placement based on a combination of two methods namely, a deterministic risk estimation for the risk assessment and a dynamic programming for optimizing the placement of surveillance cameras. That is, the enhanced efficiency of coverage is obtained by developing a prior grid assessment practice to stress on the security sensitive zones. Then, the dynamic programming algorithm operates on security quantified maps rather than uniform grids. The attained result is compared to the respective heuristic search algorithm outcomes. The overall assessment shows the reliability of the proposed methods' combinations.

Automated summary

This article explores the efficiency of optimizing the placement of surveillance cameras based on deterministic risk estimation and dynamic programming. By prioritizing grid assessment in security sensitive zones and applying dynamic programming algorithm to security quantified maps, the results are compared with heuristic search algorithm outcomes. The overall assessment shows the reliability of the proposed methods' combinations.