120 likes | 130 Views
Learn about GrabCut, a method for separating foreground and background in images using graph cut modeling. Iterated Graph Cuts ensure convergence and energy optimization for complex examples.
E N D
GrabCutInteractive Foreground Extraction using Iterated Graph CutsCarsten RotherVladimir KolmogorovAndrew BlakeMicrosoft Research Cambridge-UK
GrabCut – Interactive Foreground Extraction2 Problem Fast & Accurate ?
Foreground (source) Min Cut Background(sink) Cut: separating source and sink; Energy: collection of edges Min Cut: Global minimal enegry in polynomial time GrabCut – Interactive Foreground Extraction5 Graph Cuts modelling in images Image
Foreground (source) Min Cut Background(sink) GrabCut – Interactive Foreground Extraction5 Graph Cuts for foreground extraction Assume we know foreground is whiteand background is black
Foreground (source) Min Cut Background(sink) GrabCut – Interactive Foreground Extraction5 Graph Cuts for foreground extraction Assume we know foreground is whiteand background is black Data term = (cost of assigning label) Regularization = (cost of separating neighbors)
Foreground (source) Min Cut Background(sink) GrabCut – Interactive Foreground Extraction5 Graph Cuts for foreground extraction Assume we know foreground is whiteand background is black Data term = whiteness (cost of assigning label) Regularization = (cost of separating neighbors)
Foreground (source) Min Cut Background(sink) GrabCut – Interactive Foreground Extraction5 Graph Cuts for foreground extraction Assume we know foreground is whiteand background is black Data term = whiteness (cost of assigning label) Regularization = color match (cost of separating neighbors)
GrabCut – Interactive Foreground Extraction6 We are all set now ! ? User Initialisation Graph cuts to infer the foreground Learn foreground color model
GrabCut – Interactive Foreground Extraction6 Iterated Graph Cuts ? User Initialisation Graph cuts to infer the foreground Learn foreground color model
GrabCut – Interactive Foreground Extraction7 Iterated Graph Cuts Guaranteed toconverge 1 3 4 2 Result Energy after each Iteration
GrabCut – Interactive Foreground Extraction10 Moderately straightforward examples … GrabCut completes automatically
GrabCut – Interactive Foreground Extraction11 Difficult Examples Camouflage & Low Contrast Fine structure No telepathy Initial Rectangle InitialResult