Content based image retrieval using error diffusion block truncation coding features
1. CONTENT-BASED IMAGE RETRIEVAL USING ERROR DIFFUSION BLOCK
TRUNCATION CODING FEATURES
ABSTRACT
A new approach to index colorimages using the features extracted from the error
diffusion blocktruncation coding (EDBTC). The EDBTC produces two colorquantizes and a
bitmap image, which are further processedusing vector quantization (VQ) to generate the image
featuredescriptor. Herein two features are introduced, namely, colorhistogram feature (CHF) and
bit pattern histogram feature(BHF), to measure the similarity between a query image andthe
target image in database. The CHF and BHF are computedfrom the VQ-indexed color quantize
and VQ-indexed bitmapimage, respectively. The distance computed from CHF and BHFcan be
utilized to measure the similarity between two images.As documented in the experimental result,
the proposed indexingmethod outperforms the former block truncation coding basedimage
indexing and the other existing image retrieval schemeswith natural and textural data sets. Thus,
the proposed EDBTCis not only examined with good capability for image compressionbut also
offers an effective way to index images for the contentbasedimage retrieval system.