Dense stereo disparity maps by Dynamic Time Warp with sparse features and FPGA acceleration

Dense stereo disparity map provides a robust method on distance measurements in a 3D scene. Due to its computational complexity in nature, it has been a challenge to achieve real-time processing. In the past years, various algorithms have been studied in order to reach the goal of real-time processing. In this paper, a Dynamic Time Warp (DTW) algorithm is proposed. The DTW algorithm provides excellent results of dense stereo disparity maps in accuracy. Furthermore, an improvement on the DTW algorithm, which is the DTW with coarse quantization, is suggested to reduce the computational expense so as to meet the requirement of real-time processing. A hardware acceleration approach on an FPGA is also discussed to realize processing at a video rate.