Segmentation Of Strip Defects Image Based On Target Feature Analysis

Strip steel is one of the important steel enterprise products, its quality directly decides the quality of related terminal products. The traditional strip surface quality detection mainly depends on manual detection, which inevitably cause detection errors and missing, is difficult to meet the needs of modern production. With the development and application of machine vision and image processing technology, the steel defect detection based on image processing algorithms has become a hot study strip steel enterprises at home and abroad.At present, the classical image processing method based on the strip defect detection method mainly from the image data analysis, does not take into account the target characteristics, therefore, it is difficult to meet the integrity and detection accuracy requirements of the strip. On the basis of human visual attention mechanism, this paper proposed a method of image segmentation method based on target feature analysis. First of all, this method formats the strip significant image by feature extraction, wavelet multi-resolution filtering, central peripheral difference operation, feature integration and so on. Then, through the analysis of characteristics of strip defect image, constructs the task driven of multiple features and set up the model of visual attention mechanism based on analysis of characteristics and finish the segmentation of strip defect image. On the basis of this model, form a new algorithm. Finally, using this algorithm and the traditional detection algorithm to carry on the contrast experiment.Experimental results show that this algorithm can achieve accurate detection of various common strip defects, and the detection with less time and strong robustness, which can solve the problem of low accuracy and low efficiency in the process of detecting the traditional defects. It can provide a new method for strip production enterprises, so as to improve the quality of strip products and market competitiveness, and has broad application prospect.