Segmentation Algorithm Based On Cold-rolled Strip Surface Defects Of The Digital Filter

The background of surface image of steel rolling armor plate is very complex,some of them are common armor plate image and galvanization armor plate image etc.By reason of absence of general image segmentation theory, it is difficult to find aproper image segmentation arithmetic that is effective both to common armor plateimage and galvanization armor plate image. At the same time, most of the hugesteel-rolling factory has been improving their requirement on automatic detectingsystem for surface quality of armor plate. Proposed new image segmentationarithmetic for the surface image of steel rolling armor plate based on digital filter.After analysis the usual approach to the image defect field, proposed optimizedfilter, which can make the defect image segmentation an easier task. Furthermore,designed and realized the optimized filter, and shown the experiment result. Therobust and performance of proposed approach have been analysis and verified bytesting various images acquired from product-line. At the same time, proposed otherapproach of defect segmentation based on Gabor filer and show the result image also.Result was concluded as follow:The proposed approach using optimized filer was proved to be robust and completely satisfied defect image segmentation of steel rolling armor plate while the Gabor filer didn't result in promising performanceAs to calculation complex, the approach using Gabor filter was excellent than that using optimized filter, especially when the filer has large maskThe performance of defect image segmentation mostly depends on the size of mask. However, the performance of Gabor filer lies on the parameters specified by users.Multi-Channel optimized filer was proposed for the first time, and was used tosegment surface image of steel rolling armor plate. Discussed the prince and robustfrom the angle of both theory and realization. The defect image segmentation bycombining the way of Fourier analysis and Gabor filter result in excellent result.Therefore, the paper is significative to the field of defect image segmentation.