The on-line inspection technique of the low profile power inductor using machine vision

• Main Author: 鄭光宏
• Other Authors: Chaio-Shiung Chen
• Format: Others
• Language: zh-TW
• Published: 2004
• Online Access: http://ndltd.ncl.edu.tw/handle/92745819360354294113

碩士 === 大葉大學 === 機電自動化研究所碩士班 === 92 === This paper develops an on line inspection technique for the power inductor processing using machine vision. Image-processing techniques are also developed to identify the various defects of the power inductor, including image position, image matching, and symbol recognition. The inspection process is divided into two stages. At the first stage, we calculate the relation parameter between the measured image and the standard image to classify the surface defects. At the second stage, we recognize the symbols of the inductor and the color of copper. We extract the features of the symbols by recording the ends and nodes of the symbols and then use those features to identify the character on the inductor. More over, we RBF network to classify the color of copper. The RGB colors of the copper are firstly normalized and used as the inputs of the neural networks. We use recursive least square and back propagation methods, respectively, to training the neural networks and compare their performance. Finally, we apply the developed techniques to practical power inductor manufacturing processes to confirm the validity of the proposed method.