Skin detection using color, texture and neighborhood information
Zhi-wei JIANG,Min YAO,Jing WANG,Heng LIANG
DOI: https://doi.org/10.3969/j.issn.1006-7043.2006.z1.042
2006-01-01
Abstract:Only pixel-level color information is used in the skin detection method based on Skin Probability Map (SPM). If more other low-level cues of image are used, the skin detection accuracy can be increased further. Based on SPM method, a new skin detection method integrating color, texture, and neighborhood information is proposed. To reduce false acceptance rate, after SPM color filter, a texture filter is constructed based on texture features extracted form Gabor wavelet transform; to increase true acceptance rate, pixels passing both color and texture filters are used as seeds to implement diffusion process. Experimental results show that the proposed method can achieve better performance both in true and false acceptance than that of SPM method.