作者: Senem Velipasalar , Benjamin B. Kimia , David Mumford , Martha S. Joukowsky , Eileen L. Vote
关键词:
摘要: A heretofore unsolved problem of great archaeological importance is the automatic assembly pots made on a wheel from hundreds (or thousands) sherds found at an excavation site. An approach presented to estimation mathematical models such 3D measurements sherds. Bayesian formulated beginning with description complete set geometric parameters that determine distribution sherd measurement data. Matching fragments and aligning them geometrically into configurations based matching break-curves (curves pot surface separating fragments), estimated axis profile curve pairs for individual fragments, number features groups break-curves. Pot bottom-up maximum likelihood performance-based search. Experiments are illustrated which were broken purpose, dig located in Petra, Jordan. The performance measure can also be aposteriori probability, many other types information included, e.g., wall thickness, color, patterns surface, etc. This viewed as learning object unorganized free-form clutter, or perceptual grouping.