Motion Analysis and Research of Local Navigation System for Visual-Impaired Person Based on Improved LK Optical Flow

Abstract

For the problem of low detection accuracy and slow speed to fast motion object, when solving the basic optical flow constraint equation with the traditional algorithm, an improved optical flow algorithm based on LK optical flow algorithm has been put forward in this paper. The intensive optical flow is the biggest characteristic of the algorithm, the structure of the sampling pyramid executive optical flow, the second from bottom to date, and the second from bottom optical flow numerical multiply 2, through the double linear interpolation get the bottom of the optical flow. In the final layer iterative initialization, set specific optical flow threshold value, before a layer of iterative result for less than the response of the threshold value point, known as moving range too small point, it directly set the optical flow number to zeros and skip this point, reduce the computation time. The results shows that this improved optical flow algorithm, which can accurately analysis and forecast in the scene or particular movement of the target of the campaign mode, has the advantages of not only a high precision of motion estimation and a strong anti-interference, but also a better speed compared with the tradition optical flow algorithm.