Interested In Target Detection And Tracking Algorithm In Video Image Research

In recent years, the computer vision technology has applied broadly and quickly indifferent fields with the rapid development of computer technology and the digitalimage processing technology, like geology, astronomy, military, medical science andother fields. Moving target detection and tracing is one of the most fundamental andimportant research project in this field. The paper study the main methods in those twoparts, moving target detection and moving target tracking. In the part of moving targetdetection, the paper introduces the frame-difference method and background-differencemethod firstly, then focused on the studying of Gaussian mixture model method, andimproved this method in order to improve the accuracy of target detection. In the part oftarget tracking, study the Harris feature point detection, SUSAN feature point detectionand SIFT feature point detection. Then the paper proposes an improved Harris featurepoint detection method by building the scale space in order to improve the feature pointstability. Last the paper compared the results between the improved Harris feature pointmethod and the other methods. All the experiments were programmed in the Visual C++6.0development environment.