| The engineering ceramic has widely applied in astronautics, the nuclear industry, the petroleum chemical industry, Petrochemical industry and so on. However, the detection methods of the surface damage of the grinded ceramic imitate the metals’by now, which can’t response the damage status of the engineering ceramics effectively, so it is significant to research on detecting for the surface damage of engineering ceramics. In this paper, the time-frequency analysis has been researched to detect for the surface damage of engineering ceramics, and a set of engineering ceramic grinding surface damage detection system has been established. The main work included in the dissertation is shown as follows:1. By observing and analyzing a large number of the pictures of the surface of engineering ceramics, the description of typical damage is defined, such as pores, cracks, scratches, broken, and based on which, a set of the characteristic parameters for each damage is determined.2. Making use of the texture analysis, the texture parameters are extracted in the time-domain to determine whether the texture is existed or not. The frequency-domain filter is used to remove the grinding grain when the image has, so that the interference of the grinding texture can be eliminated3. Based on the analysis of a variety of damage characteristics and combining the pattern recognition theory, a classifier which has less damage characteristics and classification results are satisfactory, has been developed.4. Using the SD2003 industry microscope as the hardware platform and making use of Matlab7.0 GUI technology to develop the ceramic grinding surface damage detection software system, the engineering ceramics surface damage detection system is established, it’s functions including image acquisition, texture feature extraction, image pre-processing, image segmentation, feature extraction, damage evaluation etc. This system is reliable and the evaluation result is ideal. |
PDF Full Text Request