How difficult is it to recognize these objects in matlab?

Question

Image processing

There is an image:

It is necessary to determine the number of matches. I myself counted 14 matches there. But I see an option only to determine them by sulfuric heads.
This is what happened in the end after some simple manipulations.

Naturally, in this way it finds 12 objects. The
question is, do you think it is possible to separate the 2 lower matches (only at the level of coursework and matlab, and not dissertations using the edge of the opencv branch)?
I confess that I didn’t use a lot of very different methods (noise suppression, increased contrasts, erosion, etc.), because when I open it on a regular monitor, looking at it from an average distance, I don’t even see the possibility of separating them.
If it is impossible to separate them in simple ways, then which version of "counting" is preferable? For example, to calculate the average area of ​​the object, or something else?

Answer 1

Try the Hough transform . The result should be a cluster of points, the number of these clusters will be the number of matches.

Answer 2

It is not effective on sulfur heads, but to select the contours of objects and look for the number of rectangles - it will be more precise here. It is possible to apply still morphological filters what to apply.
Of the imperial methods, you can try to count the number of parallel lines.

Answer 3

Hello, you can ask your questions in the Exponent's community: https://hub.exponenta.ru/, dedicated to the use of MATLAB tools for various tasks and the exchange of experience and knowledge with authors throughout Russia.