A Real-time Motion Target Detecting and Tracking System Based on TMS320DM642

For the superior tracking performance of real-time and robustness obtained with low computational complexity, CAMShift Algorithm is of great importance in video tracking. However, it's a semi-automatic tracker, which is in need of judging and recognizing target by manual. In this paper, an automatic tracking system has been designed by the means of combining differencing-frame and CAMShift. Firstly, foreground/background is detected through sequential double-differencing-three-frame method. The CAMShift Algorithm is then used to calculate the exact location of the target and adjust the size of search window. The location information is transformed into the characters and sent via the serial port lastly, to control the tracker for maintaining the target being always visible in the scene. The algorithm has been implemented on the hardware with the core of DM642. Our experimental results show that, the system is efficient as a video tracker without human under high contrast environment.