I would like to crop photo such that only "necessary" thing left - in this case I mean text of photographed page left. I hope, that the belowe example demonstrates it clear.
I would like to do my task in python. I guess opencv library may be helpful, but what algorithms would you suggest to use? Write some own heuristics basing on image statistics, or are there some ready tools, and there is really no need for reinventing the wheel?