python - Graph Cut: create a grid graph from an image and set custom weight of edges connecting pixels -

python - Graph Cut: create a grid graph from an image and set custom weight of edges connecting pixels -

in algorithm i'm developing need utilize graph cutting images using opencv python. have define new graph image custom weight of arcs , utilize graph cutting find min cut. know there grabcut opencv function uses user-specified bounding box around object segmented find background/foreground.

here grabcut weight of edges description:

the weights of edges connecting pixels source node/end node defined probability of pixel beingness foreground/background. weights between pixels defined border info or pixel similarity. if there big difference in pixel color, border between them low weight.

i don't have find background of image, need similar opencv function possibility set custom weight of edges connecting pixels, , utilize graph cutting find min cut.

any idea?

thank you!

python algorithm opencv graph edges