Extract information from images

I am trying to extract information from the attached image of a coal sample, as for example the length and with of the fractures and the connection between them. I am using a filter and turn it into a black and white image, but have not found the right method yet, to get extract data that I can use for a statistical analysis. I have been checking the bioinformatics toolbox, which contains methods to describe the connections between biograph objects (depth first search, shortest path algorithm), but it doesn't seem to be what I am looking for. Does anyone have experience on a similar problem and could give me some useful hints?

Answer 1

[Image Analyst]

Try the Image Processing Toolbox.

[Julia]

Sure, I have been going through it all. For example using the hough transform for object analysis. I get a representation of the lines that I am looking for (see attached pictures). But my problem is how to get actual values, e.g. length and width im cm, to use them in a statistical analysis. For example I want to say in 1000 pictures I have a number of x lines with an average length of y.

[Image Analyst]

You need to spatially calibrate. See my spatial calibration demo, attached.

[Julia]

This is an awesome demo, thanks for that. In general it does the job, but I am looking for a way to automate the search!? I have been going down several paths but do not seem to find the right solution. Since I have a binary matrix I suppose it would be some algorithm counting all subsequent ones in an array of zeros..