Defects Detection of Cold-roll Steel Surface Based on MATLAB

The measure of detecting the surface edge information of cold-roll steel sheets has been investigated rely on the digital image processing toolbox of the mathematic software MATLAB, and the edge detecting experiment of surface grayscale image has been conducted on the computer. The method of detecting defects such as black flecks and scratching has been realized in the research, the different effect of operators with the disturbance of noise has been compared, the performance of the LOG´s edge detection ability various due to the change of the parameter Sigma. Conclusions have been made that LOG operator maintains the satisfying performance with the disturbance of noise, smaller Sigma is, less satisfying the smoothing ability is, which maintains more details, whereas the smoothing ability is better with loss of more details, defects detection of cold-roll steel surface based on MATLAB has a quite satisfying performance.