Angle of lines in binary image

Question

Jan Kubicek 2021-6-6

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/849405-angle-of-lines-in-binary-image

评论： Matt J 2021-6-7

Dear colleagues,

I would like to kindly ask you about computing angles in binary image. I have this binary image (and many similar), containing three binary objects. I need to fit such binary (white) objects with lines, and mainly compute angles of this lines to each object would be represented by one angle.

I tried to do that by I was not successful. I just know that MATLAB contains the inbuilt function polyfit(), which should be the way to approximate such binary objects but I cannot implement that. I would need to store these angles in some variable. For instance, when I have such binary image with three binary objects, this variable such contain three angles.

Thank you very much for the help.

.

1 个评论
显示 -1更早的评论隐藏 -1更早的评论

Matt J 2021-6-7

I just know that MATLAB contains the inbuilt function polyfit(), which should be the way to approximate such binary objects

polyfit may fail if the best fit line is perfectly vertical, or nearly so.

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Matt J 2021-6-6

1
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/849405-angle-of-lines-in-binary-image#answer_718350

编辑：Matt J 2021-6-6

Image.mat

It may be helpful to use linear2dFit from this FIle Exchange package

https://www.mathworks.com/matlabcentral/fileexchange/87584-object-oriented-tools-for-fitting-conics-and-quadrics

load Image
S=regionprops(BW,'PixelList');
imshow(BW);
hold on;
for i=1:numel(S)
   
    lfit(i)=linear2dFit(S(i).PixelList.');
    
    xy=[lfit(i).p1;lfit(i).p2];
    line(xy(:,1),xy(:,2),'Color','c','LineWidth',3);
    
end
hold off
Angles=-[lfit.angle], %line angles in degrees

Angles =
   38.2326   43.5104   55.4355

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

Answer 2

Image Analyst 2021-6-6

1
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/849405-angle-of-lines-in-binary-image#answer_718390

image.png

Try polyfit like this:

% Read in image (the posted image is RGB).
binaryImage = imread('image.png');
% Convert from color image to logical
if size(binaryImage, 3) > 1
	binaryImage = binaryImage(:,:,1) > 0;
end
imshow(binaryImage); % Display image.
axis('on', 'image'); % Display axis tick marks
% Identify separate connected components.  (Could also use bwlabel)
props = regionprops(binaryImage, 'PixelList');
for k = 1 : length(props) % For each blob...
	% Get coordinates of the points in the blob.
	x = props(k).PixelList(:, 1);
	y = props(k).PixelList(:, 2);
	% Fit a line
	coefficients = polyfit(x, y, 1);
	% Get endpoints of the line
	x1 = min(x);
	y1 = polyval(coefficients, x1);
	x2 = max(x);
	y2 = polyval(coefficients, x2);
	hold on;
	% Draw line from left most x to rightmost x.
	line([x1,x2], [y1,y2], 'Color', 'r', 'LineWidth', 3);
	% Compute angle with origin at (1,1) which is the upper left corner of the image, as usual.
	angle(k) = atan2d(y2-y1, x2-x1);
	% In case you wanted it from the other corner, the lower left corner.
	angle2(k) = atan2d(y1-y2, x2-x1);
end
% Report angles to command window.
angle
angle2