Local Harmonic Analysis and Stereo Tokens

In this paper, we discuss the problem of selecting features (tokens) that may be matched when recovering depth from stereoscopic pairs of images. A class of biologically plausible tokens, based on two-dimensional Gabor filters, is presented. It is shown that, with these tokens, one can get high qual...

Descripción completa

Detalles Bibliográficos
Autor: JOSE LUIS MARROQUIN ZALETA
Tipo de recurso: informe técnico
Estado:Versión publicada
Fecha de publicación:1988
País:México
Institución:Centro de Investigación en Matemáticas
Repositorio:Repositorio Institucional CIMAT
Idioma:inglés
OAI Identifier:oai:cimat.repositorioinstitucional.mx:1008/765
Acceso en línea:http://cimat.repositorioinstitucional.mx/jspui/handle/1008/765
Access Level:acceso abierto
Palabra clave:info:eu-repo/classification/MSC/Análisis Armonico
info:eu-repo/classification/cti/1
info:eu-repo/classification/cti/12
info:eu-repo/classification/cti/1203
info:eu-repo/classification/cti/120302
Descripción
Sumario:In this paper, we discuss the problem of selecting features (tokens) that may be matched when recovering depth from stereoscopic pairs of images. A class of biologically plausible tokens, based on two-dimensional Gabor filters, is presented. It is shown that, with these tokens, one can get high quality, dense disparity maps for images of real scenes, using a very simp!~ matching strategy. A feedback method for selecting an optimal set of filters is also presented.