基于DSP的足球机器人目标识别

Object recognition of Soccer robot based on DSP

目标识别是足球机器人视觉子系统的关键,论文对M i r o S o t中型组足球机器人视觉子系统的预处理和目标识别提出了改进的算法。基于D S P提出了改进的提取目标物体几何特征进行模式匹配的方法,实现了对复杂环境中目标物体的识别,并通过Code Composer Studio(CCS)进行了仿真测试,有较高的识别率。

Object recognition is the core of the vision subsystem in the whole system of MiroSot series Soccer-Robot. This paper proposes an improved algorithm for the vision subsystem of MiroSot competition in the middle-size league, including pretreatment of image and object recognition. Based on DSP, the geometric feature of the object images are extracted to perform pattern matching, and the object recognition in the complex environn^nt is achieved. The performance of the system has been investigated in the CCS, and the recognition shows that the system is quite efficient.