A New Algorithm for Interactive Structural Image Segmentation - Computer Science > Computer Vision and Pattern RecognitionReportar como inadecuado




A New Algorithm for Interactive Structural Image Segmentation - Computer Science > Computer Vision and Pattern Recognition - Descarga este documento en PDF. Documentación en PDF para descargar gratis. Disponible también para leer online.

Abstract: This paper proposes a novel algorithm for the problem of structural imagesegmentation through an interactive model-based approach. Interaction isexpressed in the model creation, which is done according to user traces drawnover a given input image. Both model and input are then represented by means ofattributed relational graphs derived on the fly. Appearance features are takeninto account as object attributes and structural properties are expressed asrelational attributes. To cope with possible topological differences betweenboth graphs, a new structure called the deformation graph is introduced. Thesegmentation process corresponds to finding a labelling of the input graph thatminimizes the deformations introduced in the model when it is updated withinput information. This approach has shown to be faster than other segmentationmethods, with competitive output quality. Therefore, the method solves theproblem of multiple label segmentation in an efficient way. Encouraging resultson both natural and target-specific color images, as well as examples showingthe reusability of the model, are presented and discussed.



Autor: Alexandre Noma, Ana B. V. Graciano, Luis Augusto Consularo, Roberto M. Cesar-Jr, Isabelle Bloch

Fuente: https://arxiv.org/







Documentos relacionados