The addition of openCV is great. But how do you exchange data efficiently between Tops and openCV / python ?
The method of sampling a top will be definitely to slow.
How do we manage it ? It is on the road map and new method for this purpose ?
while an easier conversion to go from TOP to DAT for openCV in python is on our radar, I’m not sure if this will ever be suitable for real-time analysis.
What i found easiest currently is to write a TOP to disk and load it back into openCV via cv2.imread().
In an earlier experiment I was also able to open a video camera from within openCV, so you could do analysis on camera streams as well.
All in all though if you need full real-time performance, the best option would be the C++ TOP or CHOP.