DocumentCode
740779
Title
Human colour skin detection in CMYK colour space
Author
Sawicki, Dariusz J. ; Miziolek, Weronika
Author_Institution
Inst. of Theor. of Electr. Eng., Warsaw Univ. of Technol., Warsaw, Poland
Volume
9
Issue
9
fYear
2015
Firstpage
751
Lastpage
757
Abstract
Fast detection of skin colour in images is an important task for computer vision and image processing. In this study, a new thresholding method for detecting skin colour has been introduced. The proposed method uses CMYK colour space, which, although not popular in image processing, especially in skin detecting algorithm, turned out to be a good choice. Its properties seem the most appropriate for the task of skin colour detection. Some experiments have been worked out using Compaq Database - a large database of skin and non-skin photos. A comparison of the new method against well-known thresholding methods in RGB, YCbCr and HSV spaces were prepared. In this comparison, the proposed method in CMYK colour space wins. This good result was further improved using a simple cascade algorithm, where two spaces RGB and CMYK were used.
Keywords
computer vision; image colour analysis; image segmentation; object detection; skin; visual databases; CMYK colour space; Compaq Database; HSV space; RGB space; YCbCr space; cascade algorithm; computer vision; human colour skin detection; image processing; thresholding method;
fLanguage
English
Journal_Title
Image Processing, IET
Publisher
iet
ISSN
1751-9659
Type
jour
DOI
10.1049/iet-ipr.2014.0859
Filename
7224064
Link To Document