【摘要】 研究基于嵌入式系统的机器视觉跟踪算法的优化,针对目前机器视觉图像处理算法存在计算量大、执行效率低及实时性较差的现状,在构建了机器视觉三维空间成像模型的基础上,从传统机器视觉图像处理算法流程入手,通过对图像处理流程的图像去噪环节和目标跟踪环节的优化设计,提出了一种面向嵌入式系统的机器视觉动态实时跟踪算法,同时对算法进行了试验,试验结果表明该算法能够自动捕捉目标物,并对目标物进行三维空间定位和动态跟踪,仿真测试结果表明了优化算法相比于传统图像处理算法,在降低了运算量的同时具有更高的执行效率和实时性,能够达到基于嵌入式系统的图像处理算法的相关要求。更多还原

【Abstract】 Algorithm optimization on robot vision tracing based on embedded system was researched in this paper. Aiming at the situation that current robot vision image processing algorithm has large calculation,low operation effective and bad real-time performance,the robot vision model of imaging in 3D space was built first,and started with the processing of image vision,the image de-noising link and target tracking link in the image processing chain were both optimized,so the dynamic tracing algorithm based on embedded system was pointed out. After experiment with the optimization algorithm,the experiment results show that this optimization algorithm can seize the target,locate in 3D space and take it into dynamic tracing,while the simulation result shows that this optimization algorithm can be high effective and better real-time performance at the same time of reducing the computational complexity. In a word,this optimization algorithm suits for the relevant requirements of robot vision image processing based on embedded system.更多还原