Histogram intersection based image retrieval technique using relevance feedback

Author

Barbu, Tudor ; Costin, Mihaela ; Ciobanu, Adrian

Author_Institution

Inst. of Comput. Sci., Romanian Acad., Iasi, Romania

fYear

2009

fDate

July 29 2009-Aug. 1 2009

Firstpage

67

Lastpage

70

Abstract

This article provides a color-based image retrieval technique for RGB image databases. Our proposed CBIR system uses the query by example approach and a relevance feedback mechanism. Feature extraction process is performed by computing a global color histogram for each image. Feature vectors are compared using the histogram intersection difference metric, and a relevance feedback mechanism is used in the retrieval process.

Keywords

content-based retrieval; feature extraction; image colour analysis; image retrieval; relevance feedback; statistical analysis; visual databases; CBIR system; RGB image database; color-based image retrieval technique; feature extraction; feature vector; global color histogram; histogram intersection difference metric; query by example approach; relevance feedback; Content based retrieval; Digital images; Feature extraction; Feedback; Histograms; Image databases; Image retrieval; Information retrieval; Sections; Spatial databases; CBIR; RGB image; color histogram; color-based image retrieval; feature vector; histogram intersection; image base; query by example; relevance feedback;

fLanguage

English

Publisher

ieee

Conference_Titel

Soft Computing Applications, 2009. SOFA '09. 3rd International Workshop on

Conference_Location

Arad

Print_ISBN

978-1-4244-5054-1

Electronic_ISBN

978-1-4244-5056-5

Type

conf

DOI

10.1109/SOFA.2009.5254878

Filename

5254878