Using Fuzzy Histograms and Distances for Color Image Retrieval
暂无分享,去创建一个
Traditionally, contentbasedimageretrieval wasbasedon the representationof the imagecontentthroughthefirstorderor second-orderdistribution of somefeaturesthoughtto be relevant for the imagedescription. Luminance, color (expressedin variouscolor representationsystems),edgestrengthandtextural featureswereintensi vely used (independentlyor jointly) during the lastdecade.We proposeto revisit theuseof color imagecontentasan image descriptorthroughtheintroductionof fuzziness,whichnaturallyarisesdueto theimprecisionof thepixel colorvalues andhumanperception.We proposetheuseof both fuzzy color histogramsandtheir correspondingfuzzy distances for theretrieval of color imageswithin variousdatabases.
[1] Vasile Buzuloiu,et al. Fuzzy Nonlinear Filtering of Color Images: A Survey , 2000 .
[2] Isabelle Bloch,et al. Fuzzy mathematical morphologies: A comparative study , 1995, Pattern Recognit..
[3] Ramesh C. Jain,et al. Similarity measures for image databases , 1995, Electronic Imaging.
[4] Isabelle Bloch,et al. On fuzzy distances and their use in image processing under imprecision , 1999, Pattern Recognit..