基于DSnet网络的零件边缘轮廓提取  

Part Edge Contour Extraction Based on DSnet Network

作　　者：于微波[1] 周旺 杨宏韬[1] YU Wei-bo;ZHOU Wang;YANG Hong-tao(School of Electrical and Electronic Engineering,Changchun University of Technology,Changchun 130012,China)

机构地区：[1]长春工业大学电气与电子工程学院,长春130012

出　　处：《组合机床与自动化加工技术》2023年第2期169-173,177,共6页Modular Machine Tool & Automatic Manufacturing Technique

基　　金：吉林省教育厅项目(JJKH20210744KJ);吉林省科技发展计划项目(20200401118GX)。

摘　　要：针对传统边缘轮廓提取算法极易受到光照强度等因素的影响,出现噪声点和零件边缘轮廓点缺失问题,提出一种基于DSnet分割网络及最外围约束算法结合的零件边缘轮廓提取算法。首先,利用DSnet分割网络对零件原始图像进行分割操作,避免影响因素干扰的同时,获得分割边界明显的零件分割图像,再根据零件分割边界像素点的特点,采用最外围约束算法进行边缘轮廓提取,得到零件边缘轮廓图像。实验结果表明:提出的方法能在光照强度不均匀等因素的干扰下准确提取出完整的、单像素的零件边缘轮廓,而且提取精度的各项指标,RMSE达到了11.18,PSNR达到了27.76,SSIM达到了0.989 3,满足边缘轮廓提取精度的要求。In view of the fact that the traditional edge contour extraction algorithm is easily affected by factors such as light intensity, noise points and part edge contour points are missing, a part edge contour extraction algorithm based on the combination of DSnet segmentation network and the outermost constraint algorithm is proposed.First, the DSnet segmentation network is used to segment the original image of the part, and while avoiding the interference of influencing factors, the segmented image of the part with obvious segmentation boundary is obtained.Get part edge contour image.The experimental results show that the proposed method can accurately extract the complete and single-pixel edge contour of the part under the interference of factors such as uneven illumination intensity, and the extraction accuracy indicators, RMSE reaches 11.18,PSNR reaches 27.76,SSIM reaches 0.989 3,which meets the requirements of edge contour extraction accuracy.

关 键 词：边缘轮廓提取 DSnet算法 最外围约束算法 

分 类 号：TH16[机械工程—机械制造及自动化] TG506[金属学及工艺—金属切削加工及机床]