Research On Welding Seam Recognition Based On Features Of Image Block

With the advantages of accurate, fast, reliable, easy to digitize, etc., visual welding seamrecognition has been the research hotspot in the fields of weld recognition in recent years, andhas great application prospect in the aspects of welding process, weld defects detection,welding seam identification, etc. In visual welding seam recognition, the influence ofillumination change and the corrosion are the key factors that hinder the recognitionperformance.In this paper, the researching and improvement of the algorithms is conduct aroundthree aspects: feature extraction, training of the classification, welding seam recognition andpositioning. In the part of feature extraction, we extract features based on image block, anduse PCA to reduce the dimension of image block features. Using the cumulative contributionrate curve as a dimension reduction reference, which choose the curve slow point as a choicefor each image block features. In classification training section, we choose the BP neuralnetwork based on LM as a classifier. In order to optimize the choice of the hidden layer nodes,we use The Comprehensive Method based on formula method and growth method todetermine the network hidden layer nodes. For the part of welding seam recognition andpositioning, according to the classification results, we reconstruct recognition image, usingthe horizontal gray value average and the minimum weld width as a positioning index,determine the weld centerline.On the basis of the above study, we build a welding seam recognition and positioningsystem on MATLAB, which can process welding seam images. At last we conductexperiments to evaluate the performance of the system in three different imaging distance anddifferent light conditions. The experimental results prove that the overlapping detection rateis90.5%, the false alarm rate is7.4%, the non-overlapping detection rate is85.9%, the false alarm rate is4.7%.