Tool Breakage Monitoring Based On Image Edge Detection

Tool is one of the most important parts in CNC machine,but because of frequent contact failure workpiece,so it's easy to breakage,led directly to the output qualified rate significantly decreased,and the machine loss will also increase,has had a huge impact on subsequent processing,hence the tool failure need timely termination of processing and replacement of cutting tools,this requires real-time monitoring of tool failure situation.The method based on image edge detection has attracted much attention because of its low cost,fast speed and high accuracy.It is necessary to adjust the threshold value flexibly in detecting tool edge to extract tool edge accurately.In this paper,the image difference method as a reference,the tool image and reference are obtained by using the technique of bokeh,and proves that the tool image and reference to figure there are sufficient gradient gap,Therefore,this paper uses the gradient difference between the tool image and the reference to set the threshold value of the edge detection algorithm.The experiment proves that many algorithms can effectively extract the tool edge while filtering out the background edge.The tool length can be calculated according to the tool edge information,and compared with the standard length,the tool breakage can be monitored.Limited by the cost of machine tools,the camera hardware parameters cannot always be used to improve the accuracy of edge detection,so the sub-pixel edge detection algorithm is adopted to improve the detection accuracy.Firstly,because of the camera diffraction effect,there is an excessive area at the edge,and a three-gray edge model is used to simulate it,and the edge error correction formula is derived.Then,in view of the Zernike moment of the commonly used an odd number of templates,analyzes the asymmetry template may lead to larger error,choosing an even number of templates can reduce errors in this case,this paper puts forward the pixel quartering to select the template to reduce the error of the subpixel edge detection,the final design experiments proved that the improved algorithm can effectively reduce the subpixel edge detection error,and calculated according to the subpixel edge information tool length error is lower than the edge pixel level.