I’ve been using OpenCV over the past couple of years for all my image processing work. It’s a really extensive and useful library, and just recently, OpenCV 3.0 was released, which included bindings for Python 3! Finally one of my last reasons to continue using Python 2 has disappeared. However, there is another image processing library that I just found out about – scikit-image. Yes, that’s the same scikit as the one from scikit-learn, the awesome machine learning library for Python that I’d linked to previously in my machine learning resources post. While I plan to continue using OpenCV, I had a read through some of the scikit-image tutorials with interest, and if you’re looking for a light-weight library for image processing that plays well with SciPy, I’d definitely check it out.
If you’re interested in learning more about image processing with OpenCV, I’m preparing a presentation for the Canberra Python Users Group using the iPython Notebook which provides an introduction to using this library. I’ll upload the notebook to GitHub and write a post linking to it in a few weeks.