Video sequence image segmentation using dynamic graph cuts

摘要

Labelling image pixels for segmentation by; UL ST="-" LIconstructing a graph representing an image, the graph comprising a set of nodes each representative of an image pixel two terminals a set of N-links each connecting a pair of nodes a set of T-links each connecting one of the terminals with one of the nodes; LIassigning a capacity to each of the N-links; LIassigning a capacity to each of the T-links; LIdetermining a first min-cut/max flow solution which partitions the nodes into two or more subsets, each containing one of the terminals; LIchanging the capacity assigned to at least one of the N-links and at least one of the T-links in response to a change in the image; LIdynamically updating the first min-cut/max flow solution to take into account the changed capacities. /UL The first min-cut/max flow solution may be determined by generating a residual graph form the graph and performing an augmenting path algorithm on the residual graph, the augmenting path algorithm generating at least one search tree. The solution is then updated by updating the residual graph and performing the augmenting path algorithm on the updated residual graph reusing the search tree.