Applying Moore's Boundary Tracing Algorithm for binary image

2 次查看（过去 30 天）

Issara Laosuwan 2016-6-8

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/288479-applying-moore-s-boundary-tracing-algorithm-for-binary-image

评论： John D'Errico 2016-6-8

Hello everyone!! I have tried to implement Moore's Boundary Tracing Algorithm for Edge Linking Operation of the binary image. This is my code:

function boundary = traceit(input)

binary = logical(input); [rows,columns] = size(binary); padded(rows+2,columns+2) = 0; padded(2:rows+1,2:columns+1) = binary;

N = circshift(padded,[0 1]); S = circshift(padded,[0 -1]); E = circshift(padded,[-1 0]); W = circshift(padded,[1 0]); boundary_image = padded - (padded + N + S + E + W == 5);

boundary_size = sum(boundary_image(:)) + 1; boundary(boundary_size,2) = 0;

for i = 1:rows for j = 1:columns if binary(i,j) == 1 break; end end if binary(i,j) == 1 break; end end

initial_entry = [j,i] + 1;

neighborhood = [-1 0; -1 -1; 0 -1; 1 -1; 1 0; 1 1; 0 1; -1 1]; exit_direction = [7 7 1 1 3 3 5 5];

for n = 1:8 c = initial_entry + neighborhood(n,:); if padded(c(2),c(1)) initial_position = c; break; end end

initial_direction = exit_direction(n);

boundary(1,:) = initial_position;

position = initial_position; direction = initial_direction; boundary_size = 1;

while true for n = circshift(1:8,[0,1-direction]) c = position + neighborhood(n,:); if padded(c(2),c(1)) position = c; break; end end

direction = exit_direction(n); boundary_size = boundary_size + 1; boundary(boundary_size,:) = position;

    if all(position == initial_position) &&...
       (direction ==  initial_direction)
            break;
    end
end

boundary = boundary - 1; end

After that I try to apply this function for my target image. I have perform the following conversion on the image before using the function: RGB - Gray Scale - Binary - Canny Edge Detector

im = imread('shape.png'); img = rgb2gray(im); BW = im2bw(img,0.5); e = edge(BW,'canny'); imshow(e,[])

e = traceit(out); figure imshow(out)

After I have computed, it does not return with any error, but the algorithm also doesn't work in the manner I expected neither. Please kindly give me some advice.

Thank you very much!!!