Dopasowanie chmur punktów do powierzchni dla potrzeb przetwarzania i analizy obrazów medycznych oraz systemów wizyjnych

This paper addresses a problem of automatic point clouds matching in context of training data construction in Active Shape Model algorithm. An algorithm was applied to CT data segmentation for radiotherapy planning. Proposed methodology consists of 3 steps: point clouds generation, rigid matching (rotation, translation) based on ICP and local corrections of points location. Algorithm was tested and validated using distance between points in data after matching. Point clouds matching for medical image processing, analysis and vision systems.