curve detection in a binary image
4 ビュー (過去 30 日間)
古いコメントを表示
Hello all, below is one of the canny edge detected breast image from a dataset of 20 breast images

The curves I would like to detect are shown below in red ( sincere apologies for a messy representation, I used paint to point out the curves)

I cannot approach this problem considering the nipples because not all the edge detected images show definite nipples, also not every breast is of the same size and shape so i cant create a fixed mask either. Is there any other way to approach this and get the above marked curves? I have attatched the original image and the code within.
clc;
clear;
close all;
a=rgb2gray(imread('FLIR0541.jpg'));
en= imsharpen(a,'Radius',4,'Amount',2);
ed=edge(en,'canny',0.1);
bw1=bwareaopen(ed,10);
bw=imdilate(bw1,true(3));
b=imfill(bw, 'holes');
figure;
imshow(b);
Any help is much appreciated. Thank you in advance.
2 件のコメント
KSSV
2020 年 3 月 31 日
If the location of your required curves are always more or less in the same location, you can play with indices and get what you want..?
回答 (0 件)
参考
製品
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!