A Machine Vision Approach for Small Ceramic Tubes Detection

A method which computer vision technology is used to detect defect of small ceramic tubes is put forward, and the automatic testing system is developed by technologies of computer, computer vision, machine-electron and so on, and automatic detecting and classifying of small ceramic tubes are realized. The system comprises five parts which are the machine of automatically delivering and picking, PC, control unit, lamp-house, image collecting and processing. In testing, first, small ceramic tube running, and simultaneously, shooting and storing the tube images by CCD video camera; and secondly, image is preprocessed, segmented, feature extracted and sorted; finally, the inferior is eliminated by the manipulator. Taking example for the semitransparent ceramic tube that external diameter is 8 millimeter and inside diameter 6 millimeter, the testing method of small caliber and thin ceramic tube is researched by the automatic system. The outcome of the experimentation accords with the fact, this method is simple and feasible, and is of high efficiency, dependability, and high automation.