论文部分内容阅读
提出了基于高速数字信号处理器TMS320C6415的图像识别跟踪系统,并给出了成像跟踪系统的原理图和结构图,重点论述了采集系统和图像处理系统。在软件上采用矩不变阈值分割和基于梯度的自适应阈值分割提取目标,采用投影特征的质心跟踪算法及基于线性微分拟合的记忆外推跟踪技术跟踪目标。实验结果表明:该系统跟踪稳定可靠,完全满足实时性要求。
An image recognition and tracking system based on high speed digital signal processor TMS320C6415 is proposed. The principle diagram and structure diagram of imaging tracking system are given. The acquisition system and image processing system are mainly discussed. In the software, Moment Invariant Threshold Segmentation and Gradient-based Adaptive Threshold Segmentation are used to extract the target. The centroid tracking algorithm of projection features and memory extrapolation tracking based on linear differential fitting are used to track the target. The experimental results show that the system is stable and reliable, and fully meets the real-time requirements.