Appearance Of Capacitor And The Study Of The Image Processing Algorithm Of The Detection System

Capacitor is one of the electronic components which are widely used in electronic equipments. In order to realize capacitor’s automatic detection, using digital image processing technology, the image processing algorithms of the capacitor’s appearance measurement and analysis system are investigated, which has important practical significance.The overall design scheme of the detection system is designed, which is mainly composed a light, a CCD camera, an image acquisition card, the image processing module and the module which can distinguish defective products. The image processing is the core of the paper which focuses on the algorithm’s selection and design of the key function modules including image filtering, image segmentation, image recognition and so on.The adaptive median filter is adopted for image to remove noise. According to the different characteristics and testing requirements of lead, bottom and profile images, the corresponding segmentation algorithm is designed. Combining the gradient image and morphological operations for the lead image’s segmentation and get the lead gray image which background removed is obtained. Using the morphological operations, the gradient of the image, the Ostu method and the contour tacking algorithm, the segmentation of the bottom of the image is completed. Adopting k-means clustering algorithm, and combining with Canny operator edge detection for the side image’s edge segmentation and restoration are achieved. Experimental results show that the effect is good.To detect the different capacitor defects, different features for extraction and recognition algorithm are designed. Through the analysis of difference between the two lead gray images filled and unfilled, the corrosion of lead is judged. The circular degree is applied to judge the defect that the incision is not round. The image is changed from RGB model to HSI model. Based on the analysis on the H component and S component, the bottom surface’s condition of the liquid leakage is distinguished. The shape defect of the side image is identified according to the rectangular features. Through the segmentation of the segmented image by K-mean clustering, supplemented by color feature analysis, the judgment of the white defect is realized.At the platform of Matlab R2007a, the simulation study of the image processing algorithm are completed, and a capacitor appearance image detect software is developed which laid the foundation for the capacitor appearance detection system.