损伤零件缺陷区域的三维形貌检测对于激光机器人加工过程的智能化具有重要作用。系统实现了激光加工中损伤零件的缺陷识别,并提出损伤零件缺陷边界检测的高精度算法。利用基于点特征面积累加的法矢法识别出缺陷边界点,生成初始缺陷边界,针对初始缺陷边界具有不连续、多分支、锯齿状边界等问题,提出了三步优化处理方法,最终得到封闭、高精度的损伤零件的缺陷边界。系统运行处理速度快,精度高,可完全满足工业零件智能化激光再制造的要求。 The detection of the three-dimensional topography of the defect area of ​​the damaged part plays an important role in the intelligent process of the laser robot. The system realizes the defect identification of the damaged parts in the laser processing, and puts forward the high precision algorithm of the damage detection of the defect boundary. The boundary points of defects are identified by using the normal vector method based on the accumulative points of characteristic points of the points, and the initial defect boundaries are generated. Aiming at the problems of the discontinuous, multi-branch and jagged boundaries of the initial defect boundaries, a three-step optimization approach is proposed. Finally, High-precision damage parts of the defect boundary. System running fast, high precision, fully meet the industrial laser parts intelligent laser remanufacturing requirements.