论文部分内容阅读
提出了一种机器人“手-眼”视觉分步动态目标定位方法,实验中保持摄像机图像平面与工作台平面平行,并对摄像机的拍摄高度进行设定,最终将目标定位在图像的中心位置。在目标类型识别过程中,采用BP神经网络技术对目标表面字符进行识别,通过识别的字符判定目标类别。最终由字符信息及目标定位信息引导末端执行器完成对目标的准确分选。
This paper proposes a robot vision system with “hand-eye” visual step-by-step dynamic target positioning. In the experiment, keeping the camera image plane parallel to the workbench plane and setting the shooting height of the camera, the target is finally located in the center of the image position. In the target type identification process, BP neural network technology is used to identify the target surface characters, and the target type is determined by the recognized characters. Ultimately by the character information and target location information to guide the end-effector to complete the target accurately.