I can utilize the matching result of Hungarian algorithm to connect corresponding clusters using B-spline method directly, rather than first interpolate any two candidate clusters. What’s more, the pairwise results of Hungarian algorithm can help me to infer the where occlusion occurs, i.e. it tells me the positions or regions where occluders are. Thus with a number of these Hungarian results, we can get a final occluder map which defines where occluder exists.
Archive for the ‘Mics’ Category
2010-10-30 Leave a comment