Tiny visual feature extraction of random changing weld

To random variation characteristics of the welding process of thin steel lap weld,Proposing a methods based on the energy distribution of the visual featuresdetection and extraction,Using pseudo-color enhancement algorithms to get energy distribution, Effectively suppressed splash, smoke, noise and other transient weak energy during welding.Then proposing a differential search algorithm to achieve the accurate extraction of structured light stripe skeleton,gaining the image feature points.Then using a random sample consensus algorithm to analysis probabilistic of image feature historical data,Thus achiving accurately fitted local model of weld features, fulfiling accurate prediction of feature points. Experimental results show that the proposed method is effective and weld visual feature extraction satisfactory.