Reputation: 3783
I have a need to find the leading edge in a machine vision application. This edge can vary slightly in shape and can sometimes have lower contrast than an edge further inside the part we are measuring.
I would like to find a general algorithm I can use to find the leading edge in a region I specify, searching in a direction I specify. Lastly, this must always be a positive edge. That is, it transitions from a dark region to a lighter region.
To be clear, I am not looking for a edge detection algorithm, like Canny. I want help on the general approach to selecting the right edges in my image that fit within an ROI and match the polarity constraints.
Here is an example of what I am talking about:
Upvotes: 1
Views: 51
Reputation: 1258
The Canny feature detection of opencv
does the job. Extract features, and later you can filter them however you want (e.g. the transistion, direction, etc.)
Upvotes: 1