Title :
A probabilistic method of skin detection
Author :
Mitra, Subhasish
Author_Institution :
Dept. of Electr. Eng., Jadavpur Univ., Kolkata, India
Abstract :
This work presents a novel technique for detection of skin color. This method is based on probabilistic modeling of skin pixel testing, which has produced highly encouraging results. This paper also introduces the method of using trained neural networks for fitting probability distributions. A comparison with other popular methods has been provided in this paper.
Keywords :
curve fitting; image colour analysis; learning (artificial intelligence); neural nets; object detection; skin; statistical distributions; fitting probability distribution; probabilistic modeling; skin color detection; skin pixel testing; trained neural networks method; Hysteresis; Image color analysis; Neural networks; Probabilistic logic; Probability distribution; Skin; Bayes´ Rule; Kolmogorov-Smirnov Test; color models; hysteresis threshold; morphological reconstruction; neural networks; skin detection;
Conference_Titel :
Image Information Processing (ICIIP), 2013 IEEE Second International Conference on
Conference_Location :
Shimla
Print_ISBN :
978-1-4673-6099-9
DOI :
10.1109/ICIIP.2013.6707558