Steel slab surface defect detection using texture analysis

Hot rolled steel slabs during their manufacturing process encounter several faults which are visible on the surface.

Using texture analysis we differentiate the defects on the steel slab surface images.

Here for the analysis we use the processing using LBP algorithm, from that matrix we use Convolution for spot detection, edge detection and level detection.

The code starts by taking an image from dataset given as follows;

I=imread('cameraman.tif');

defect(I)

gives three spot, edge, level detections.