论文部分内容阅读
传统算法存在识别精度较低,实时性较差的问题。为了提高识别精度,提出基于视觉分析的高精度实时激光加工路径识别算法,首先利用合理的点运算分析法提高激光加工目标物体图像的对比度;其次通过迭代阈值分割法与二值图像闭运算提取激光加工目标物体图像特征;然后对提取的激光加工目标物体路径信息进行滤波处理;再利用适应度函数对激光加工目标物体路径进行排序;最后通过视觉分析法按排序结果进行依次识别,选取出激光加工目标物体的最优路径。实验结果分析证明,利用该算法能够实现激光加工目标物体路径的高精度识别,且实时性强。
The traditional algorithm has the problem of low recognition accuracy and poor real-time performance. In order to improve the recognition accuracy, a high-precision real-time laser path recognition algorithm based on visual analysis is proposed. Firstly, the contrast of the laser processing target object image is improved by a reasonable point operation analysis method. Secondly, the laser is extracted by iterative threshold segmentation and binary image closing algorithm Processing the image features of the target object; filtering the extracted path information of the laser processing target object; using the fitness function to sort the path of the laser processing target object; and finally, sorting by visual analysis according to the sorting result, selecting the laser processing The optimal path of the target object. Experimental results show that the algorithm can realize high-precision identification of laser processing target object path and real-time performance.