Segmentação de imagens tomograficas usando contornos deformaveis com segmentos de custo minimo

AUTOR(ES)

Elisangela de Araujo Rodrigues

DATA DE PUBLICAÇÃO

2000

RESUMO

Tomographic images contain important information about the anatomy (or function) of three-dimensional structures of the human boby. Image segmentation is the operation that "extracts" the structures of interest from the tomographic images for a given aplication. Deformable contours are among the most investigated methods for image segmentation. Two traditional problems in such an approach are the convergence and the stabilization of the deformable model over a desired object boundary. This work presents an overview about deformable contours and proposes an efficient solution to the convergence and stabilization problems. Our solution essentially splits the contour in minimum-cost segments which snap onto the object boundary when the deformable contour gets close to it. The results of the work are illustrated with respect to image segmentation tasks in medicine

ASSUNTO(S)

tomografia processamento de imagens

Documentos Relacionados

• Segmentação do lúmen em imagens de IOCT usando Fuzzy Connectedness e Reconstrução Binária Morfológica
• Precisão de medidas lineares em imagens tomograficas
• Problemas de fluxo de custo minimo em redes de custo concavo
• Formulações e algoritmos sequenciais e paralelos para o problema da árvore geradora de custo mínimo com restrição de grau mínimo
• Segmentação fuzzy de imagens e vídeos