融合改进Canny与RCF的齿轮零件边缘检测算法  

Edge detection algorithm of gear parts based on improved Canny and RCF

作　　者：李思雨 范必双[1] 杨涯文 朱之健 Li Siyu;Fan Bishuang;Yang Yawen;Zhu Zhijian(School of Electrical and Information Engineering,Changsha University of Science&Technology,Changsha 410114,China)

机构地区：[1]长沙理工大学电气与信息工程学院,长沙410114

出　　处：《现代计算机》2024年第18期34-39,共6页Modern Computer

摘　　要：随着智能制造的发展,齿轮作为工业常用零件,其边缘检测对于确保零件质量具有关键意义。传统的Canny边缘检测算法对弱边缘信息丢失较为严重,不能满足高精度的测量要求。为了提高测量精度,提出了一种基于改进Canny与RCF融合的方法。算法先通过RCF网络对齿轮边缘进行检测,然后将Canny算法得出的结果与改进RCF网络检出的关键边缘进行融合得到最终边缘。实验结果表明,改进的算法对于齿轮的边缘检测的精度和抗噪性有更好的效果。With the development of intelligent manufacturing,gears,as commonly used industrial components,their edge detection is crucial for ensuring part quality.The traditional Canny edge detection algorithm severely loses weak edge information,failing to meet the requirements for high‑precision measurement.To enhance measurement accuracy,a method based on the fusion of improved Canny and RCF has been proposed.The algorithm first detects gear edges using the RCF network,then fuses the results from the Canny algorithm with key edges detected by the improved RCF network to obtain the final edges.Experimental results show that the improved algorithm achieves better accuracy and noise resistance in gear edge detection.

关 键 词：齿轮零件 边缘检测 改进RCF算法 改进Canny算法 融合算法 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] TH132.41[自动化与计算机技术—计算机科学与技术]