Performance Analysis of Color Space for Optimum Skin Color Detection

In the development of Human computer interface applications information related to skin color is extensively utilized. Reason being skin is the biggest part of human being. Human skin color shows resemblance with non-skin colored material like wall paint, wood etc. So an efficient human computer interface (HCI) system is required to be designed which can distinguish between them. The default color space obtained from imaging device is RGB and processing with this color space takes more time as comparison to other color spaces, hence color space transformation is required. Thus in this paper system performance is evaluated on the basis of processing time and computational complexity. It is verified that less complex structure of application reduces the processing time too. Using two different color space schemes computational time for processing a fixed number of frames of video signal in a constant environment are evaluated and on the basis of that a better color scheme method is proposed.