edge detection : Code sobel

hello,someone correct me this program please.
I=imread('lena.jpg');
I=rgb2gray(I);
[i j]=size(I);
I=double(I)/255;
level=0.5;
%carte de gradient
MH= [1 2 1; 0 0 0;-1 -2 -1];
GH = conv2 (I, MH);
figure(1),imshow(GH), title('Contours horizontal');
MV= [-1 0 1; -2 0 2; -1 0 1];
GV = conv2 (I, MV);
figure(2),imshow(GV), title('Contours vertical');
G=sqrt(GH.*GH + GV.*GV);
G=im2bw(G,level);
figure(3),imshow(G), title('Contours global');
%direction du gradient
eq=imdivide(GH,GV);
eq=atan(eq);
eq= (eq*180)/pi;
figure(4),imshow(eq),title ('direction gradient');
%seuillage et suppression des non maximum locaux
for i = 1 : size(eq,1)
j = 1 : size(eq,2)
if level > eq(i,j)
eq(i,j) = 0;
if level < eq(i,j)
eq(i,j) = 255;
end
end
end
eq = im2bw(eq,level);
figure(5), imshow(eq)

6 个评论

1-deletion of local non-maxima is not correct because i didn't see any different between it and the result of (contours global).
2-the exercice is (Calculate and display a gradient with normalized values ​​between 0 and 255) so i want know if i should always convert the values of image in double and do the devision ( I=double/255)
3- i can do the threshold after calculate the direction of gradient or after magnitude
thanks for help
Try imgradient() and imgradientxy()
i dont want to work with function
Why not? Is it your homework assignment? If so, tag it as homework.

请先登录,再进行评论。

回答(0 个)

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by