The thesis shows the development of DSP-based ¡§human head-tracker¡¨ system. It uses CCD camera to capture images, and detects in the image sequence. When someone interrupts, the system will lock on his head and shows the locked image on the LCD screen. The Head-tracker system includes three sub-systems : ¡§Motion Detector¡¨, ¡§Ellipse Algorism¡¨, and ¡§Visual Probability Data Association Filter¡¨. From the results of experiment, it can meet the expectation and gain good performance and robustness.