how to calculate Jaccard index using segmented image with contour of original iamge in matlab -
i have segmented image , got contour of original image using command imcontour(original_image,1) , have used code calculate jaccard index in matlab.i new in matlab please me.here code have used
function [jaccardidx,jaccarddist] = jaccard_coefficient(img_orig,img_seg) img_orig = logical(img_orig); img_seg = logical(img_seg); if ~islogical(img_orig) error('image must in logical format'); end if ~islogical(img_seg) error('image must in logical format'); end inter_image = img_orig & img_seg; union_image = img_orig | img_seg; jaccardidx = sum(inter_image(:))/sum(union_image(:)); jaccarddist = 1 - jaccardidx;
but , @ time of intersection of 2 images, error comming matrix dimensions must agree , not able resolve it.thanks in advance please me.
after reading comments understand reason failure fact using matrices different dimensions: ground truth image bigger segmentation image([121 167] vs [120 161]).
in general, these images should have same size in order calculate ir jaccard index. you'll need check why did different sizes in first place.
if 100% sure it's not bug, can perform resizing of segmentation image according the size of ground truth:
segimage = imresize(segimage,size(origimage));
Comments
Post a Comment