Active visual continuous seam tracking based on adaptive feature detection and particle filter tracking

• Rong Fan, Peng Zhang, Fengyun Guo, Jie Rong, and Xupeng Lian
• received 01/30/2024; accepted 04/15/2024; posted 04/16/2024; Doc. ID 520506
• Abstract: Welding seam tracking based on online programming is the future trend of intelligent production. However, most of the existing image processing methods have certain limitations in the adaptability, accuracy and robustness of weld feature point detection. The on-line welding method of gas metal arc welding (GMAW) based on active vision sensing is studied in this paper. Steger sub-pixel detection method is used to guarantee the accuracy of feature point extraction, and self-adaptive search window and self-adaptive slope extraction are proposed on this basis, which has certain robustness and universality for continuous weld detection. When arc light and other serious interference makes it difficult to obtain weld information, particle filter is used to make the best prediction of weld position. Finally, the welding robot platform based on laser vision sensing was built to test various continuous welds of butt weld, fillet weld and lap weld. Through the detection test of weld point on laser stripe image and the tracking performance test during welding tracking, experimental results show that the detection speed is 27ms, the accuracy of detection and tracking can respectively reach 0.80pixel and 0.78mm, which meet the requirements of weld detection and tracking.