An intelligent tracking and monitoring system based on SoPC

This paper proposes a way to implement moving target tracking and monitoring system based on System-on-Programmable-Chip (SoPC). This approach can improve the flexibility and reduce the over-reliance on host computers of the traditional embedded tracking and monitoring devices. An adaptive moving target extraction algorithm based on Gaussian model is introduced in this article and implemented in hardware-oriented ways, which improves the performance of the system. We construct a dual-core SoPC, respectively, to achieve moving target tracking and Linux-based real time video network transmissions. When operating, the Pan-Tilt-Zoom (PTZ) can drive the camera to track the moving target in sight. After testing, when the system operates at 100MHz with hardware acceleration, the time consumed of processing a video frame is about 37.32 milliseconds, and the speed can be accelerated for 41.70 times.