武汉理工大学数字图像处理英文课件 Chapter2（4）.pptVIP

Region Labeling of Steel Grains Clean Image Skeletonize and remove spurs Find Grain Boundaries and Label Grains * Load in a sample image showing a microscope image of steel. load imdemos steel imshow(steel) The steel image is thresholded at two different threshold values bw_70 = steel70; bw_210 = steel210; bw_70 bw_210 Points which were black in the image with the low threshold are used to select white objects in the negative of the image with the high threshold. This effectively removes the minor regions which appeared in the second thresholded image. [r,c] = find(bw_70==0); bw_clean = bwselect(~bw_210, c, r, 8); imshow(bw_clean) [i,j] = find(X) returns the row and column indices of the nonzero entries in the matrix X. This is often used with sparse matrices. BW2 = bwselect(BW1,c,r,n) returns a binary image containing the objects that overlap the pixel (r,c). r and c can be scalars or equal-length vectors. If r and c are vectors, BW2 contains the sets of objects overlapping with any of the pixels (r(k),c(k)). n can have a value of either 4 or 8 (the default), where 4 specifies 4-connected objects and 8 specifies 8-connected objects. Objects are connected sets of on pixels (i.e., pixels having a value of 1). The BWMORPH function is used to skeletonize the result of the minor region removal. bw_skel = bwmorph(bw_clean, skel, 6); imshow(bw_skel); BWMORPH is applied again to prune the spur pixels from the skeletonized image. bw_pruned = bwmorph(bw_skel, spur, 8); imshow(bw_pruned); The grain boundary map is the logical NOT of the pruned image. grain_boundaries = ~bw_pruned; imshow(grain_boundaries); [labeled,N] = bwlabel(grain_boundaries,4); colored = label2rgb(labeled); imshow(colored); The BWLABEL command is used to label different objects in the binary result. The objects are displayed with different colors from a jet colormap. *