Edge detection using matlab

close all;
clc;
img=imread('download.jpg');
B=rgb2gray(img);
subplot(2,2,1)
imshow(B)
pause(2)
I=double(B);

for i=1:size(I,1)-2
for j=1:size(I,2)-2
%Sobel mask for x-direction:
mx=((2*I(i+2,j+1)+I(i+2,j)+I(i+2,j+2))-(2*I(i,j+1)+I(i,j)+I(i,j+2)));
%Sobel mask for y-direction:
my=((2*I(i+1,j+2)+I(i,j+2)+I(i+2,j+2))-(2*I(i+1,j)+I(i,j)+I(i+2,j)));

B(i,j)=sqrt(mx.^2+my.^2);
end
end
subplot(2,2,2)
imshow(B); title('Sobel gradient');
pause(2)
%Define a threshold value


Thresh=100;
B=max(B,Thresh);
B(B==round(Thresh))=0;
B=uint8(B);

subplot(2,2,3)
imshow(~B);title('Edge detected Image');

Im using this code for detect the edges in my image and it works fine ...

My question how to plot outpot line on the input image using least squares fitting ???