Hi, I have an imaginary bot that walks on a corridor and takes pictures of the hall, have to make the bot to understand where is the end of the hall. Main.m is the main file.
I have a picture of the corridor. I used the edge function to detect the edges of the image of the corridor. I blurred the image to avoid noise. Used different edge functions (for comparison) and do not know if I put the correct formula for the Mean Square Error to compare them to show which is better. I have to compare the edge function's results between them, to show that edge_prewit is the cleanest and has the best result, i want to show this with MSE and with other different methods, for example luminance (luminosity).
Now I have to do the bot to understand where is the end of the corridor with other words to detect the square (the shape of the square with white outline).
After detecting the square I have to trace the lines from the corners of the square to make Matlab (or my imaginary bot) to see the whole corridor (not just the square).
I have attached all the files i use in program.zip, and now simply paste the codes here:
Main.m
originalRGB = imread('coridor1.jpg');
h = fspecial('gaussian',4,8); %blurr the image
I=imfilter(originalRGB,h);
figure(1), imshow(originalRGB), figure(2), imshow(I); %Show the blurred image
imwrite(I,'coridor3.jpg','jpg')
figure(3), I3=edgecolor_prewit('coridor3.jpg'), imshow(I3) %edge prewit
figure(4), I4=edgecolor('coridor3.jpg'), imshow(I4) %edge canny
figure(5), I5=edgecolor_sobel('coridor3.jpg'), imshow(I5) %edge sobel
%subplot(2,2,1), imshow(originalRGB);title('Original image');
%subplot(2,2,2), imshow(I); title('Blurred image');
% h = fspecial('gaussian',6,8);
% I4=imfilter(I3,h);
% figure(4), imshow(I4)
[m,n] = size('coridor1.jpg');
error = originalRGB-I3;
MSE = (sum(sum(error.*error)))/(m*n)
error2 = originalRGB-I4;
MSE2 = (sum(sum(error2.*error2)))/(m*n)
error3 = originalRGB-I5;
MSE3 = (sum(sum(error3.*error3)))/(m*n)
edgecolor.m
%Finding edges of a color image
%Usage edgecolor('female.jpg');
function S=edgecolor(nm);
img=imread(nm);
[x y z]=size(img);
if z==1
rslt=edge(img,'canny');
elseif z==3
img1=rgb2ycbcr(img);
dx1=edge(img1(:,:,1),'canny');
dx1=(dx1*255);
img2(:,:,1)=dx1;
img2(:,:,2)=img1(:,:,2);
img2(:,:,3)=img1(:,:,3);
rslt=ycbcr2rgb(uint8(img2));
end
S=rslt;
% Q=imread('coridor1.jpg');
% figure, imshow(Q)
% figure, imshow(R)
edgecolor_prewit.m
%Finding edges of a color image
%Usage edgecolor('female.jpg');
function S=edgecolor(nm);
img=imread(nm);
[x y z]=size(img);
if z==1
rslt=edge(img,'canny');
elseif z==3
img1=rgb2ycbcr(img);
dx1=edge(img1(:,:,1),'canny');
dx1=(dx1*255);
img2(:,:,1)=dx1;
img2(:,:,2)=img1(:,:,2);
img2(:,:,3)=img1(:,:,3);
rslt=ycbcr2rgb(uint8(img2));
end
S=rslt;
% Q=imread('coridor1.jpg');
% figure, imshow(Q)
% figure, imshow(R)
edgecolo_sobel.m
function R=edgecolor(nm);
img=imread(nm);
[x y z]=size(img);
if z==1
rslt=edge(img,'log');
elseif z==3
img1=rgb2ycbcr(img);
dx1=edge(img1(:,:,1),'log');
dx1=(dx1*255);
img2(:,:,1)=dx1;
img2(:,:,2)=img1(:,:,2);
img2(:,:,3)=img1(:,:,3);
rslt=ycbcr2rgb(uint8(img2));
end
R=rslt;
% Q=imread('coridor1.jpg');
% figure, imshow(R)
% figure, imshow(Q)
and the image
Thanks a lot in advance! PLEASE PLEASE PLEASE HEEELP!!!
