BDBComp
Parceria:
SBC
RECUPERAÇÃO DE IMAGENS DE FACES HUMANAS BASEADA EM CONTEÚDO UTILIZANDO WAVELETS E P.C.A. (PRINCIPAL COMPONENT ANALYSIS)

Marcelo Franceschi de BianchiAdilson Gonzaga

This work describes a novel and efficient algorithm for contentbasedimage retrieval based on the discrete Wavelet Transform(DWT) and Principal Component Analysis (PCA), togetherwith inputs drawn from the Euclidian distance operator, acommon criterion for distance measurement. The former is usedto produce a signature vector from the query input image, acompressed and codified vector that holds the key features ofthe original image, and the latter is used to make projection ofthe images onto proper subspaces. Interestingly, the tests statethat, for every particular query, the worse the frequencyresponse of the wavelet filter, the better the classification,where the accuracy the algorithm up to 98,61 %. The system´sinput consists of a query image and its output corresponds tothe most similar image found in the data base, according to thedistance criterion adopted.

http://www.lbd.dcc.ufmg.br/colecoes/wvc/2005/002.pdf

Caso o link acima esteja inválido, faça uma busca pelo texto completo na Web: Buscar na Web

Biblioteca Digital Brasileira de Computação - Contato: bdbcomp@lbd.dcc.ufmg.br
     Mantida por:
LBD