On-Site Tool Wear Detection Based On High Precision Computer Vision

Tool wear is one of the key factors affecting the quality and machining efficiency. With the development of science and technology, on-process tool wear detection has become more important, in order to meet the higher requirement of the surface quality and economy. In comparison with the traditional methods, machine vision technology is a simple and high precision quick solution for the tool wear detection, without contact, and distortion. However, existing computer vision tool wear detection systems still have problems, like off-line testing, test elements of a single, non-automated and low detect efficiency.Based on its evolution, the following exploratory works have been done in order to solve the problems exist in the current research of tool wear detection technology:First, combining manipulator with six degrees of freedom realized on-site tool wear detection based on computer vision, and proposed five measurements elements based on wear characteristics of micro-diameter cutter. The tool is placed on the spindle, computer vision system clamped by robot moves to three measurement position, to obtain the tool image.Second, the realization of automatic detection. Including automatic image acquisition and automatic image processing. Automatic image acquisition section realized movement system automation, auto-focus, automatic light control and magnification automatic control, automatic image processing part realized automatic image pre-processing and feature extraction of tool image, the entire process is automated detection in2min.Finally, the analysis and compensation of measurement error. Based on the analysis of systematic error sources and common error compensation technology, proposed error compensation method based on the length of the experiment corrected measurement accuracy of±3μm up to±1μm, and gives the area compensation algorithm based on the length compensation.