This a C++ reimplementation of the papaer efficient graph-based image segmentation. I just change the code in https://github.com/congve1/SaliencyRC from python to C++. Besides, I use the count sort method in the sorting step,so my code can process a 1200 X 800 image within 2s and a 1900 X 1500 iamge at about 36s.
There are many strings "heheXX" in my code, they are my debug flag. :)
All you need is to change the image path and config the opencv environment.
-C++
-opencv3.4+
submit my learning in HC.and repair the bug in graph-based-segmentation. some middle result in HC
quantize:
smooth:
The moment the sgementation like.
Now I can go on for RC.