Does Colorspace Transformation Make Any Difference on Skin Detection?

  • Authors:
  • Min C. Shin;Kyong I. Chang;Leonid V. Tsap

  • Affiliations:
  • -;-;-

  • Venue:
  • WACV '02 Proceedings of the Sixth IEEE Workshop on Applications of Computer Vision
  • Year:
  • 2002

Quantified Score

Hi-index 0.00

Visualization

Abstract

Skin detection is an important process in many of computervision algorithms. It usually is a process that starts ata pixel-level, and that involves a pre-process of colorspacetransformation followed by a classification process. A colorspacetransformation is assumed to increase separabilitybetween skin and non-skin classes, to increase similarityamong different skin tones, and to bring a robust performanceunder varying illumination conditions, without anysound reasonings. In this work, we examine if the colorspacetransformation does bring those benefits by measuringfour separability measurements on a large dataset of805 images with different skin tones and illumination. Surprisingresults indicate that most of the colorspace transformationsdo not bring the benefits which have been assumed.