Recognition of the type of welding joint based on line structured-light vision

Abstract

To recognize the type of welding joint is an essential precondition for extracting features of weld seam and guiding robot tracking seam automatically. A method based on a line laser structured-light vision for recognizing the type of welding joint is studied in this paper. Images of welding joint captured by camera are preprocessed firstly for noise reduction and enhancement with wavelet transform, and the reconstructed images are converted to binary ones using appropriate thresholds. Then some features of binary images are further extracted and formed feature vectors which are input into a PNN classifier for classification. Combined with the position relationship of laser and camera, four types of welding joint are eventually recognized. Experimental results show that, this method has a high recognition rate.