摘要
提出了一种用于机器人手臂实时视觉伺服的运动图像实时处理的方法。这种方法采用全局遗传算法 /局部遗传算法以及未经处理的原始图像对一个已知模型的运动目标物体进行识别。在 GA处理的过程中 ,适应度函数的计算是基于表面—条纹模型作为目标物体模型 ,并且同时采用全局遗传算法和局部遗传算法来识别物体的形状及检测位置与方向 ,提高了 GA的收敛速度 ,并得到了良好的识别结果。还采用一个三角形的物体为目标 ,对所提出的方法进行了仿真试验 。
This paper presents a motion image processing technique for a manipulator real time visual servoing which utilizes global search feature of a genetic algorithm (GA) and a local research technique of GA and also unprocessed gray scale image which is called here raw image, in order to perform recognition of a known target object. Also in GA process, the computation of fitness function is based on the configuration of an object model designated as surface strips model. The global GA is utilized together with local GA in order to recognize the object shape and detect position and orientation simultaneously and to increase the GA's convergence speed so as to provide faster and better recognition results. Computer simulation experiments of recognition and tracking of a triangle object has been done and the results demonstrate that the proposed scene recognition method is very effective.
出处
《仪器仪表学报》
EI
CAS
CSCD
北大核心
2002年第z1期170-172,共3页
Chinese Journal of Scientific Instrument