Image Recognition Based on Neural Network Method for Automated Welding Machine

• Main Authors: Chia-Lang Lu, 盧家朗
• Other Authors: Seng-Chi Chen
• Format: Others
• Language: zh-TW
• Published: 2008
• Online Access: http://ndltd.ncl.edu.tw/handle/44503424919131656712

碩士 === 大葉大學 === 電機工程學系 === 96 === This thesis is mainly aimed at the combination of motion control and image processing techniques for automated welding application. Image recognition is based on a modified Hopfield neural network method. The proposed method in this study has the features of parallel processing, high error tolerance and learning-recalling intelligence capacity. To demonstrate the feasibility and effectiveness of the proposed method, a Cartesian robot arm on which a CCD is rigidly mounted performs precision position control over work pieces. Experimental results indicate that the manipulator is successfully applied to the automated welding work. Moreover, the proposed method significantly enhances reliability, quality and efficiency.