How to calculate the length of the longest crack in this image automatically ?

How to highlight those cracks in this image and calculate their lengths?

[Image Analyst]

It's complicated by the fact that the really long crack crosses over dark circular regions. Do you consider the crack to end when it hits a round dark region? Or do you want it to continue through it? Please annotate the image with red lines showing exactly where each crack is.

[Jayanti]

Hi,

Please refer to the below links on calculating crack length which might be useful for you:

https://in.mathworks.com/matlabcentral/answers/2076311-determining-crack-length-from-images

https://www.mathworks.com/matlabcentral/answers/196886-image-processing-for-crack-detection-and-length-estimation

Additionally, you can refer to the below link for calculating crack area, width, and length:

https://www.mathworks.com/matlabcentral/fileexchange/176438-crack-width-length-and-area-calculation-and-visualization

Answer 1

[KALYAN ACHARJYA]

img = imread('image_crack.png');
% Enhance contrast 
enhancedImg = imadjust(grayImg);
% Apply Gaussian filter to reduce noise
smoothedImg = imgaussfilt(enhancedImg, 2);
% Binarize the image using a manually set threshold
binaryImg = imbinarize(smoothedImg,0.32);
%figure, imshow(binaryImg);
%get the highest blob
BW = bwareafilt(~binaryImg,1);
figure,imshow(BW);

Appooximate Only: Now, attempt to use a skeleton-based approach or approximating triangles to determine spatial coordinates and calculate the distances.

[wd w]

KALYAN ACHARJYA, thank you very much!