The Crucial Technical Study Of On-Line Detection Based On CCD In Digital Curve Grinding Field

Nowadays, there are so less studies of complicated curve-face grinding technology and equipments in china, especially in multi-varietal, flexible, multi-functional grinding machine field, that it can not catch the needs of quick development in mould manufacture and complicated tool manufacture field. Therefore, it is significant to study high-quality, multi-functional, on-line detection grinding technology. This thesis comes from the program: the technology and equipment of precise curve grinding machine based on on-line CCD image detection (021111125), which was sponsored by Shanghai Science Committee. In order to catch the needs of precise grinding technology in on-line detection curve grinding field, the study selects three main targets, which are as follows: developing the technique to detect the wearing of grinding wheel and compensate it dynamically in the process of grinding; developing the precise tool position technique for grinding wheel and workpiece; developing the on-line image detection technique to measure the dimension and the quality of workpiece. The whole thesis bases on three key technologies: the technology of dynamic compensation of wearing of grinding wheel, the technology of cross-cursor tool position, and the technology of on-line detection of workpiece. Via experimentation, all the three key technologies are basically finished. Based on actual environment, the author has studied how to improve the quality of images and the precision of image detection, developed the method to diminish the image noises by using hardware and software, developed the method of four-point calibration and its detection of precision, and established the ways of edge crawl in different occasions by comparing the two ways of edge search. From the experiments, using the dynamic compensation of wearing of grinding wheel can make the whole contour precision of workpiece within 5μm, and precision of cross-curse tool position and on-line detection can achieve 2μm.In this thesis the author has independently finished the following items: the selection of image processing in practical application, and using hardware to diminish the spark; enhancing the quality of image calibration and developing the measurement to check it.; putting the image system into MD9040 by using VC++; developing the tool position system by cross-cursor technique.