Application of Optical Flow Field for Intelligent Tracking System

In order to make the data responding to the edges of moving objects expand to the area, whose gray level is flat, it needs to keep iterative numbers large enough on gradient-based optical flow algorithm. The traditional approaches need a large amount of calculation, but results accuracy is low. The gradient number accuracy and reliability will directly determine the result of optical flow algorithm. In the hope of improving numerical reliability of image gradient data, we have used four methods to manage gradient number effectively, which are Hessian matrix distinguishing, Gaussian filtering standard deviation amending, mean model amending and multi-image comparing. The four algorithms mentioned above were applied to tracking contours and segmentation of car images on the traffic scene. From the experimental results, it is shown that both the calculation convergence speed and accuracy of our methods have greatly improved compared with traditional algorithms, the feasibility and validity of those methods were confirmed.