With the rapid advancement of object tracking technology, Multi-Target Multi-Camera Tracking (MTMCT) technology has been widely used in the fields of video surveillance and intelligent traffic. However, most research efforts have focused on improving the accuracy of vehicle tracking, neglecting its application in real-time traffic monitoring scenarios. During task execution, the complexity of the vehicle environment and the high computational demands of the selected models result in significant time consumption, hindering real-time deployment. To address this problem, this paper proposes a parallel processing real-time multi-camera vehicle tracking system that balances detection speed and accuracy. This system performs frame-by-frame parallel detection and tracking through multi-cameras, transmitting the obtained tracking data to the multi-camera vehicle clustering stage. Through complex feature matching and spatiotemporal information association, it achieves global vehicle trajectory association. By integrating these strategies, the system can support scalable real-time vehicle tracking application in complex traffic environment. This paper demonstrates the effectiveness of the proposed system through experiments, highlighting its potential for widespread adoption in intelligent transportation system.