Oriented image foresting transform segmentation with connectivity constraints

Nenhuma Miniatura disponível
Data
2016
Autores
Mansilla, Lucy A. C.
Miranda, Paulo A. V.
Cappabianco, Fabio A. M. [UNIFESP]
Orientadores
Tipo
Trabalho apresentado em evento
Título da Revista
ISSN da Revista
Título de Volume
Resumo
A new algorithm, named Connected Oriented Image Foresting Transform (COIFT), is proposed, which provides global optimum solutions according to a graph-cut measure, subject to high-level boundary constraints. COEFT incorporates the connectivity constraint in the Oriented Image Foresting Transform (OIFT), ensuring the generation of connected objects, and can also handle simultaneously the boundary polarity. While the connectivity constraint usually leads to NP hard problems in other frameworks, such as the min-cut/max-flow algorithm, COIFT conserves the low complexity of the OIFT algorithm. Experiments show that COEFT can improve the segmentation of thin and elongated objects, for the same amount of user interaction.
Descrição
Citação
2016 Ieee International Conference On Image Processing (ICIP). New york, p. 2554-2558, 2016.