How do I highlight very faint lines from an image?

I am trying to check snapshots from a live video source for the existence of lines. I was hoping to do this with the edge() function, however, the real lines are too faint and instead the granularity of the image is highlighted. Is there a way to alter the image to make the lines stand out more? Altering the actual imaging setup is not an option.

I have attached an example photo. Some of the faint lines that I would like to highlight can be seen in the lower right part of the image.

Answer 1

[Peter O]

Have you tried the Hough transform?

https://www.mathworks.com/help/images/ref/houghlines.html

[Jason Johnson]

It’s my understanding that in order to use the hough transform functions you first have to use the edge function to find the lines in the image, which currently I am unable to do.

[Peter O]

None of the edge options work?

How regular in thickness and vertical are you expecting the lines to be? Is there a chance you can use a bandpass operation to filter out some of the high frequency noise before you run it through edge?

[Jason Johnson]

They all pick up the granularity of the image instead of the lines. The lines should be vertical and somewhat regular in thickness so a bandpass filter could work. How do you propose I apply one to the image?

[Peter O]

I don't have much experience with desgining image filters, but a few other contributors do. Does Image Analyst's suggestion in the post below help? It looks like you might be able to pick out the peaks in the spectral plot:

https://www.mathworks.com/matlabcentral/answers/494415-bandpass-filter-applied-to-an-image

If it the spectrum is obscured in the full image, perhaps try cropping the test image to the portions containing the lines to get the pixel frequency.

[Jason Johnson]

Thank you, I will try to make use of this to build a filter.