The Study Of The Image Inspecting Technique And The Chip Shape Recognition Technique

The control degree of the cutting process is the key for determining whether the automation manufacturing system could run normally, so the monitoring technique concerned is essential. But the traditional monitoring methods have define limitation and can not adapt to the different conditions. In this paper based on image analysis theory the chip shape recognition and tool wear inspecting were analyzed and studied.In this paper firstly the image analysis system making use of chip shape recognition and tool wear inspecting were established, it was made of industry computer, image collecting card of QP300, pickup camera of WV-BP330CCD and microscope. This affords foundation for the image analysis application to cutting process. The image analysis system was demarcated in condition of fixed distance and magnifying diploid microscope by experiments and the demarcated results was obtained.The chip shapes in the course of cutting processing was carefully analyzed and studied and put forward feature expanded and radbas function neural network algorithms for recognition chip shape, more than experiments program was developed. feature expanded algorithm was the method which makes use of subsistent feature and rotates these feature and XOR operation, the new feature which was man-made created was as recognizing basis. Radbas function neural network method makes use of self-learning function of neural network and automatic adjusted tentatively value, so self-adapted capacity of the algorithms was strength. The results showed that the algorithms took on these characteristics of training speed rapidly and short time and good adapted capacity for different data convene.Tool wear image characteristic was studied at tool wear inspecting aspect and put forward '十 window' automatic edge detection algorithm and resolved the problem of tool wear image edge extraction. The algorithms and programs calculating area, perimeter, width and length for tool wear image were developed for tool wear image. The better results were gained by experiments and improved edge orientation precision and geometry measure precision. Finally aiming at the sawtooth chip discussed the calculating problem of sawtooth chip measure and put forward edge draft method applying to angle and length's calculation.