Image retrieval by color semantics with incomplete knowledge

Region-based image retrieval using integrated color, shape, and location index

“…Many approaches exist for color quantization that include vector quantization, clustering [21,47] and neural networks [9]. The entire RGB color space can be represented using a smaller set of color categories that are perceptual to humans.…”

“…The entire RGB color space can be represented using a smaller set of color categories that are perceptual to humans. A possible theoretical justification for this is provided by Corridoni et al [47], who present a system supporting image retrieval by high-level chromatic contents, which are meant as sensations that color accordances generate on the observer. Images are archived by describing the spatial arrangement of regions with homogeneous chromatic attributes.…”

“…Following [21,43,47], and as per the recent MPEG-7 standardization directives for visual descriptors [42], we too use a color quantization in RGB space using 25 perceptual color categories to segment images based on dominant colors. This is justified since these 25 colors are sufficient to clearly distinguish all the colored regions in our chosen domain of image database.…”

Customising a qualitative colour description for adaptability and usability

“…Lammens [4] presented a computational model for colour perception and colour-naming based on the CIE XYZ, CIE L * a * b and neuropsychophysical (NPP) colour spaces. Corridoni et al [5] presented a model for colour-naming based on the hue saturation and lightness (HSL) colour space and also introduced some semantic connotations, such as warm/cold or light/dark colours. Mojsilovic [6] presented a computational model for colour categorisation and naming and extraction of colour composition based on the CIE lab and HSL colour spaces.…”

“…Therefore, the HSL colour space is well suited to be divided into intervals of values corresponding to colour names, and the semantic labels associated with these names in order to refer to the richness (saturation) or the brightness of the colour (luminance) are also intuitive [15]. Previous approaches also chose the HSL colour model for their studies [3,5,6].…”