Optimizing the camera configurations impacts the performance of the video surveillance applications. Where, proper camera placement reduces the total cost and increases the surveillance efficiency. Various methods are used to optimize the coverage such as greedy search and linear programming, hence the typical cost function for optimizing the camera placement focuses on obtaining the maximum coverage regardless of the area significance or the camera capabilities. This work proposes a novel cost function for camera placement problem. The proposed approach models the camera vision capability based on the task to be performed. The model represents the significance of the monitored area by means of risk maps. Then the coverage optimization is performed based on the area significance modeling results and the sensor capability. The outcomes show the applicability of the proposed cost function in various scenarios. © 2018 IEEE.