Based On The Texture Characteristics Of Disk Surface Grain Defect Detection

Surface inspection plays an important role in industrial production of product quality control. The surface marks defects of magnetic material in the process seriously affect its quality. Detection of these defects is not efficiency and susceptible to interference by human.There are five non-destructive testing methods mostly used in industrial production which are eddy current testing, radiographic testing, penetrant testing, magnetic particle testing and ultrasonic testing. These methods have their own limitations such as too expensive, too slow and detection process is cumbersome to meet the needs of real-time detection in industrial production.In order to achieve real-time and robust detection of surface marks defects, we proposed a method in this paper. The picture of the surface of disk is got by the computer industrial vision systems. This method which based on texture feature makes various defects to be unity by enhancing the defective parts and inhibiting the background parts after image pre-processing. Then we detect surface marks defects by designing a suitable mask. The method we proposed in this paper achieved the detection of surface marks defects.The experimental results shown that it can be done in real-time and is useful for a variety of defects.