Performance evaluation of content-based image retrieval on networks / Avaliação do desempenho de recuperação de imagens médicas baseada em conteúdo em redes de computadores

AUTOR(ES)
DATA DE PUBLICAÇÃO

2008

RESUMO

The field of information retrieval based on visual contents, has increased in importance due to its great deal of available material, including digital images and videos, which has been shared and disseminated mainly via internet. Thus, new patterns of consumption, handling, and exploration of digital materials have been developed using proper organization and indexation. The considerable increase in the digital medical image database, as well as the emerging technology and the retrieval tools based on the contents of these images, has contributed substantially to medicine practice. Additionally, associated to the picture archiving and communication systems (PACS), this advancement has positively influenced the education process and has aided diagnoses.The main objective of this work is to verify the performance of a content-based image retrieval system running on computer networks, testing the computational times involved for access on a high-speed optical (KyaTera) network and on the internet. Wavetels families were also evaluated as texture extractors and the results were compared in medical images, considering the recovery efficiency.The results obtained herein may serve as a basis for the development of image recovery programs applied to the medical sciences area, and as indicative of traffic speed on computer networks installed in hospitals, connecting them to clinics and other infrastructures in the medical field.

ASSUNTO(S)

redes de computadores medical images imagens médicas cbir cbir wavelets wavelets network

ACESSO AO ARTIGO

Documentos Relacionados