抄録
An algorithm for tracking the position of a moving object in a naturally moving image is proposed. In this algorithm, the first step is to simply identify the objcct. Then, the object is segmented from the back ground image by the local histogram growing method. The histograms in the perceptual color system (hue, saturation, brightness) are insensitive to video-camera movement and rotation, as well as deformation, shrinkage, and enlargement of the object. Last, the position of the object is detected by the optimal histogram matching method. Real-time experiments have verified the effectiveness of this technique.
