Dynamic programming based disparity estimation circuit design for 3D image system

In this paper, we present a method of reconstruct depth map with dynamic programming. By binocular camera, we get the information to calculate disparity map and find optimal path to obtain dense depth map. In order to achieve real-time system, we simplify the algorithm to reduce hardware complexity and use parallel hardware architecture to accelerate calculation. Finally, we implement a digital chip to get 640×480 resolution of depth map image for fast 3D depth map.