The extraction accuracy of intersecting line weld affects the accuracy of industrial welding,which is an important step of weld technology and one of the key issues in industrial research.In order to solve the problems existing in the traditional method,the intersection line weld extraction algorithm based on convolution neural network and bilateral filtering is proposed.Firstly,the weld image of the intersecting line is grayed and enhanced by the contour wave transformation method;Secondly,the weld image is de-noised by bilateral filtering method;Finally,the intersecting line weld seam is extracted through the full convolution neural network.The experimental results show that the proposed method has higher accuracy,faster speed and better overall application effect.
关键词
图像去噪处理/卷积神经网络/图像增强处理/双边滤波/相贯线焊缝提取
Key words
Image denoising/convolution neural network/Image enhancement processing/Bilateral filtering/extraction of intersecting line weld
维普
万方数据