China Mechanical Engineering ›› 2026, Vol. 37 ›› Issue (6): 1426-1441.DOI: 10.3969/j.issn.1004-132X.2026.06.016

Previous Articles    

Research Progresses on Tool Wear State Detection Based on Machine Vision

MENG Boyang(), LI Zhongjie, XIANG Fuxing, LI Song, ZHOU Jiaqi, YUE Caixu   

  1. Key Laboratory of Advanced Manufacturing and Intelligent Technology,Ministry of Education,Harbin University of Science and Technology,Harbin,150080
  • Received:2025-07-18 Online:2026-06-25 Published:2026-07-17
  • Contact: MENG Boyang

基于机器视觉的刀具磨损状态检测研究进展

孟博洋(), 李众杰, 向福星, 李松, 周佳琪, 岳彩旭   

  1. 哈尔滨理工大学先进制造智能化技术教育部重点实验室, 哈尔滨, 150080
  • 通讯作者: 孟博洋
  • 作者简介:孟博洋*(通信作者),男,1991年生,博士、讲师。研究方向为智能机床数控系统设计、数据驱动的刀具磨损状态智能监测技术、基于数字孪生的加工过程建模及加工质量优化等。E-mail:bymeng@hrbust.edu.cn
  • 基金资助:
    国产数控系统应用示范工程(HN工程)(2023ZY01076-11);黑龙江省重大科技成果转化项目(CG23012);黑龙江省省属高等学校基本科研业务费(2021-KYYWF-0759)

Abstract:

Key technologies and research advancements concerning machine vision for tool wear condition monitoring were systematically reviewed. Starting from detection objects, correlations between wear features and conditions, as well as extraction methods, were discussed. Typical monitoring systems, including offline, on-machine static, and dynamic monitoring systems were outlined. At the level of detection methods, detection processes and key methods based on traditional image processing were systematically summarized, and applications of artificial intelligence technologies in these detection fields were elaborated. Advantages and limitations of various visual detection technologies were summarized, and future key research directions were projected. Leaps from two-dimensional to high-precision three-dimensional dynamic detections, upgrades from single-vision to multi-sensor fusion anti-interference systems, and implementations of generalized algorithms and edge deployments from laboratories to production lines were proposed to be achieved, aiming to provide systematic references for related researches and applications.

Key words: tool wear, machine vision, image acquisition, image preprocessing, state recognition

摘要:

系统综述了机器视觉在刀具磨损状态检测中的关键技术与研究进展。从检测对象入手,探讨了各检测对象中磨损特征与状态之间的关联以及提取方法。概述了包括离线、在机静态与动态监测在内的典型监测系统。梳理了基于传统图像处理的检测流程及关键方法,阐述了人工智能技术在该检测领域中的应用。归纳总结了各类视觉检测技术的优势及局限性,并对未来重点研究方向进行了展望,提出应实现从二维到高精度三维动态检测的跨越、从单一视觉到多传感器融合抗干扰体系的升级,以及从实验室到产线的泛化算法与边缘部署落地应用,旨在为相关研究与应用提供系统性参考。

关键词: 刀具磨损, 机器视觉, 图像采集, 图像预处理, 状态识别

CLC Number: