摘要
The characteristics of X-ray testing image are analyzed and improved competitive fuzzy edge detection is described. This algorithm takes into account maximizing objective function to estimate the edge intensity at first. Then, according to the new edge patterns, learning vector quantization neural network is applied to each edge pixel according to its assigned class. At last the thinning algorithms is run to get the one-pixel wide edge image. Experimental results show that the proposed method can better improve the efficiency of weld seam image processing.
The characteristics of X-ray testing image are analyzed and improved competitive fuzzy edge detection is described. This algorithm takes into account maximizing objective function to estimate the edge intensity at first. Then, according to the new edge patterns, learning vector quantization neural network is applied to each edge pixel according to its assigned class. At last the thinning algorithms is run to get the one-pixel wide edge image. Experimental results show that the proposed method can better improve the efficiency of weld seam image processing.
