Multi-Scale Texture Feature Representation for Content-Based Image Database Retrieval

Details Multi-Scale Texture Feature Representation for Content-Based Image Database Retrieval Multi-Scale Texture Feature Representation for Content-Based Image Database Retrieval

Sep 2004

adsabs.harvard.edu/cgi-bin/nph-data_query?bibcode=2004esasp.553e...3b&link_type=abstract

Proceedings of ESA-EUSC 2004 - Theory and Applications of Knowledge-Driven Image Information Mining with Focus on Earth Observa

Statistics

Applications

Gabor Wavelet Transform, Weighted Histogram, Luminance, Chrominance, Achromatic Texture

Scientific paper

Recent systems for content-based image retrieval (CBIR) employmulti-scaleimagefilteringtechniques suitable for texture analysis. Although they compare favourably to alternative techniques that do not employ convolution filters, these multi-scale CBIR systems perform rather poorly when compared to human observers. This seems not only due to the peculiar ability of humans to infer visual features and semantic meanings from images based on prior knowledge, but also to the similarity inaccuracy introduced by: i) the feature representation (i.e., the image characteristic signature extraction), which is intrinsically non-injective, and ii) the similarity measure, whose selection depends on the set of features. This work reports on new developments in feature extraction, feature representation and distance measure selection for content-based image retrieval.

Affiliated with

Also associated with

No associations

Rating

Multi-Scale Texture Feature Representation for Content-Based Image Database Retrieval does not yet have a rating. At this time, there are no reviews or comments for this scientific paper.

If you have personal experience with Multi-Scale Texture Feature Representation for Content-Based Image Database Retrieval, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Multi-Scale Texture Feature Representation for Content-Based Image Database Retrieval will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFWR-SCP-O-1064340