Weld Seam Edge Extraction Algorithm Based on Beamlet Transform

We present a novel algorithm based on Beamlet Transform for extracting seam edges from noisy welding images. Taking into account of some special characteristics of welding image processing, we introduce an orientation-thresholding step as well as a two-scan method to the standard beamlet-based line detection algorithm. Experiments are conducted to extract edge features from noisy welding images at different scales and to test the anti-noising performance of our algorithm. The result of experiments show that our algorithm is capable of directly extracting weld seam edges from highly noisy welding images without any pre-processing or post-processing steps, thus is characterized by its high efficiency and its prominent anti-noising performance. The orientation threshold can not only reduce the calculation load, expedite its running speed, but also it can help to improve the anti-noising performance of the algorithm. The two-scan method is particularly helpful when coping with low SNR welding images.