Cutting Tool On-Machine Detection Based On Computer Vision

With the improvement of science and technology and social production,the requirements become much higher for the workpiece quality and productivity.In the process of machining,the state of the cutting tool directly determines the quality of the workpiece.In order to avoid the scrap of workpiece and the damage to machine tools,it is necessary to monitor the state of cutting tool.By the measuring of cutting tool,its accurate shape can be achieved and can be used to optimize the cutters tracks,realizing precision machining.In order to meet the requirement of on-machine monitoring and measuring the state of cutting tool,an image capturing and processing system have been established.The investigation of autofocusing,tool wear monitoring and measuring geometry parameter of cutting tool were presented in this paper.1.According to the demand of the on-machine detection system,the appropriate camera,lens and light source were selected to build an image capturing system.Comparing with sensitivity of several focusing evaluation functions,we selected Brenner gradient function to realize autofocusing,combining the movement of axis of ultraprecision machining machine tool.Besides,with coarse adjustment before fine-tuning,the process of focusing can be accelerated,capturing images with high quality.2.Researches about shape matching method and gray matching method were analyzed in three aspects,which are principle,stability and efficiency.Combining pyramid algorithm in shape matching,the cutting tool wear can be monitored rapidly.3.An algorithm was proposed for the cutting tool detection and compared with the algorithm of circle arc fitting in efficiency,stability and precision.The result showed that the self-designed algorithm's stability was better and it could accurately measure the cutting tool with notch.4.As for the measurement of the cutting tool beyond visual field,a stitching algorithm basing on contour was proposed to ensure the accuracy of detection.Contour recognition,affine transformation,and edge fusion technology were adopted to promise the accuracy of the cutting tool profile stitching.Finally,the precision of the stitching algorithm was verified in the experiment.