基于DM642的运动目标自动跟踪实验平台研制

摘要

为测试运动目标自动检测与跟踪算法在DSP系统中的运行性能，以及加强大学生创新能力的培养目标，设计了1套以DM642硬件为核心的运动目标自动跟踪实验平台，实验平台的软件系统采用模块化设计。介绍了基于背景减除的运动目标检测和改进CamShift跟踪算法的原理，并给出了运动目标自动检测和跟踪的实验结果。实验结果表明，该实验平台能够对设计的运动目标跟踪算法进行实时的运行，为跟踪算法的实际应用提供1个方便的测试平台。%In order to evaluate algorithmes performance for the moving object automatic detection and tracking application in the DSP system, and to improve the innovation ability and motivation of undergraduate students, a DM642 based tracking system in the Lab is developed. The software system of the experimental platform is designed using models, the principles of object detection based on subbackground and improved CamShift tacking algorithms is introduced in detail, and provided the experimental results of the moving objects automatic detection and tracking. Experimental results show that the tracking algorithms can run correctly on this experimental platform, and provide a testing platform for the practice applications of tracking algorithm.