matlab使用opencv （轮廓提取+直线拟合）

I =imread('1.jpg');%读取图片
I1 = double(I1);
[Gx,Gy]=gradient(I1);
G=sqrt(Gx.*Gx+Gy.*Gy)/255;%%求图像梯度

I2 = rgb2gray(G);
figure,imshow(G);

I2(I2<30/255)=0;%%%二值化
I2(I2>=30/255)=1;

U= findContours(I2);%轮廓提取
[m,n] = size(U);%得到轮廓数量
aa = ones(n,4);
for i=1:n
d = U(1,i);
[ii,jj] = size(d{1,1});
jj
if jj<10%过小的轮廓不要

continue;
end
%%%%%%直线拟合-start%%%%%%%%%%%%%

aa(i,:)= fitLine(d{1,1});
%%%%%%直线拟合-end%%%%%%%%%%%%%
end