Just starting to take a look through the Touchlib source and how it uses openCV, and I’m wondering why Touchlib seems to implement its own blob detection algo, and not using the one provided by openCV (BlobTracker class).  Or I may be wrong and not reading the code correctly - still getting my head around some of this.

I think it is just another layer of abstraction, but I haven’t really dug into it yet… I did look into a little of using python’s ctypes for interfacing with OpenCV.