Title :
Retrieval of iris dominant color using intra-palette color merging
Author :
Lodin, Adrian ; Rusu, Corneliu
Author_Institution :
Fac. of Electron., Telecommun. & Inf. Technol., Tech. Univ. of Cluj-Napoca, Cluj-Napoca, Romania
fDate :
June 30 2011-July 1 2011
Abstract :
Identifying dominant colors prominent in a visual scenery is very important since human visual system primarily uses them for perception. When dealing with eye iris images it is important to correctly distinguish dominant colors from the iris surface. In this paper we present an automatic dominant color detection method which allow us to calculate the similarities from the iris surface. Using the intra-palette color merging technique unique colors are tested for similarity by comparing the Euclidian distance between them with a similarity marker. The colors marked as similar are merged together to form a new weighted color. The process repeats until there are no similar colors but only dominant ones. The experimental results claim the efficiency of the method.
Keywords :
geometry; image colour analysis; image retrieval; iris recognition; object detection; Euclidian distance; automatic dominant color detection method; human visual system; intra-palette color merging technique; iris dominant color retrieval; Histograms; Humans; Image color analysis; Iris; Merging; Pixel; Visualization;
Conference_Titel :
Signals, Circuits and Systems (ISSCS), 2011 10th International Symposium on
Conference_Location :
lasi
Print_ISBN :
978-1-61284-944-7
DOI :
10.1109/ISSCS.2011.5978697