Online tool wear condition monitoring using binocular vision

摘要

$1School of Mechanical Engineering, Xi'an Jiaotorig University, Xi'an, China$2Xi'an Jiaotong University, Xi'an, China$3School of Mechanical Electronic and Information Engineering, China University of Mining and Technology, Beijing$4Xi'an Jiaotong University, Xi'an, China$5Key Laboratory of Education Ministry for Modern Design and Rotor-Bearing System, Xi'an Jiaotong University, Xi'an 710049, China$6School of Mechanical Electronic and Information Engineering, China University of Mining and Technology (Beijing), Beijing 100083, China; ;The real-time detection of tool wear condition plays an important role in the safety of machine operation and improving the quality of products. According to the requirements for image-based tool wear detection, an online tool wear monitoring system was designed in this paper using the binocular vision method, including the hardware and software. The proposed algorithm, by use of the speeded-up robust features (SURF) algorithm, epipolar geometry theory and block matching method, could skip the complex steps of camera calibration. The convenience and flexibility of this system has made it possible to conduct online tool wear detection in the future. Online experiments were conducted to obtain a disparity map of the tool wear images, which could reflect the tool wear condition. The results showed that the proposed method, with an accuracy of 15.6 μm, could meet the accuracy requirements for online tool wear monitoring.