How to extract location of a feature of interest from image backgroud of similar pixel intensity?

I have a large set of images in which I need to extract the location of a particle that has an insert. Since the insert is a straight, steel wire, I am tempted to use edge detection. However I fear that it may not be the best strategy to go with. As the particles in the system move around, the shape and size of the wire insert also appear to change. Could I kindly get a few ideas on how to extract location of the wire insert? Raw Images are attached.

Answer 1

[Image Analyst]

There are lots of edges in those images so I don't know how well they'd work. This might be a good situation for a deep learning approach, like with segment.

[Humair Nadeem]

I understand. I was thinking more along the lines of more traditional approaches. Apart from edge detetion and machine learning are there any other strategies that I could explore? For edge detection my strategy would have been to initially blur the image and then try to reduce the false positives. But as you said I am not sure if that would be enough. Is there a method to only capture certain shaped edges (for example could one ignore all curved/circular edges)?

[Image Analyst]

You could try template matching with normalized cross correlation. I'm attaching a demo.

[Humair Nadeem]

Thanks Image Analyst. i will try this and let you know how it went. Thanks again!