A visual analytical method for evaluating tool flank wear volumes of micro-milling cutters with AKAZE features matching: A preliminary study

作     者：Zhang, Yu Gao, Shuaishuai Duan, Xianyin Zhu, Kunpeng 

作者机构：Wuhan Business Univ Sch Mech & Elect Engn Wuhan 430056 Hubei Peoples R China JiangHan Univ Sch Artificial Intelligence Wuhan 430056 Hubei Peoples R China Wuhan Univ Sci & Technol Sch Machinery & Automat Wuhan 430081 Peoples R China Wuhan Univ Sci & Technol Key Lab Met Equipment & Control Technol Minist Educ Wuhan 430081 Hubei Peoples R China Chinese Acad Sci Inst Adv Mfg Technol Hefei Inst Phys Sci Lab Precis Mfg Changzhou 213164 Jiangsu Peoples R China 

出 版 物：《WEAR》 (Wear)

年 卷 期：2025年第564卷

核心收录：

学科分类：08[工学] 0805[工学-材料科学与工程（可授工学、理学学位）] 0802[工学-机械工程] 

基　　金：National Natural Science Founda-tion of China 

主　　题：Micro-milling Tool flank wear Computer vision and image processing Image registration AKAZE feature matching Hough transformation 

摘      要：Tool wear is one of the most important factors restricting the development and application of micro-milling technology. In order to accurately evaluate the tool wear features in the micro-milling process, a visual analytical approach is proposed with multiple advanced image processing methods. Firstly, A theoretical model of tool wear volume considering the end concave angle is proposed, which is more descriptive of practical situations than the traditional one-dimensional parameters, the tool flank wear length. Secondly, the tool flank wear areas are accurately obtained of image segmentation with the Canny gradient operator, extraction with corner feature detection, and registration with Accelerated-KAZE (AKAZE) feature matching. Finally, the features of the tool flank wear are evaluated of correcting the integration areas with the Hough transformation, determining the parameters with scale conversion, and verifying the areas and volumes of the tool flank wear with experimental data and key quality metrics, Mean-Squared Error (MSE) and Multi-Scale Structural Similarity (MSSSIM). The results demonstrate that the method presented effectively characterises the tool flank wear, giving a suitable approach for computer vision and image feature processing of tool wear monitoring during machining.