Content-based image retrieval (CBIR) has become one of the most active rese
arch areas in the past few years. Most of the attention from the research h
as been focused on indexing techniques based on global feature distribution
s. However, these global distributions have limited discriminating power be
cause they are unable to capture local image information. The use of intere
st points in content-based image retrieval allow image index to represent l
ocal properties of the image. Classic corner detectors can be used for this
purpose. However, they have drawbacks when applied to various natural imag
es for image retrieval, because visual features need not be corners and cor
ners may gather in small regions. In this paper, we present a salient point
detector. The detector is based on wavelet transform to detect global vari
ations as well as local ones. The wavelet-based salient points are evaluate
d for image retrieval with a retrieval system using color and texture featu
res. The results show that salient points with Gabor feature perform better
than the other point detectors from the literature and the randomly chosen
points. Significant improvements are achieved in terms of retrieval accura
cy, computational complexity when compared to the global feature approaches
. (C) 2001 SPIE and IS&T.