基于负曲率极值点的零件识别与检测技术研究

Study on Parts Recognition and Checking Based on Negative Curvature Minima

在机器人柔性自动装配中,零件的识别与定位是一项极为重要的技术。提出了一种新的基于点到弦的距离累积算法,以实现零件轮廓曲线上负曲率极值点的检测,并在此基础上利用短切规则实现目标轮廓的分解与特征参数的提取。首先,用平面几何区域的面积来近似平面点到弦的距离累积积分,得出了距离累积积分与曲线曲率之间的关系,利用距离累积积分来检测平面离散曲线的负曲率极值点。然后,利用短切规则对零件轮廓进行分解,获取零件的几何特征,实现零件的识别与检测。试验和仿真结果说明了该方法的有效性。

In robot flexible assembling processing, parts recognition and checking is a very important technique. A new point-to-chord distance accumulation for searching negative curvature minimum points of the parts contour curve was proposed, and the contour was decomposed by the negative curvature minimum points combined the short-cut rule, Firstly, the area of the planar geometry region was used to approximate the distance accumulation integral, and the relationship between the distance accumulations and the curvatures was deduced. Then the information of the distance accumulation curve was exploited to detect the negative curvature minima （NCM） of the object contour. Through the NCM points, parts contour was decomposed with the short-cut rule, Then the parts can be recognized and checked by its geometric features. The experiments and simulation results show the effectiveness of the proposed method.