Automatic Detection of Fine Hairs in Skin Aging Analysis

Image segmentation of skin texture is an important step in the computer aided analysis of skin aging. Poor segmentation results are obtained if fine hairs in skin images are neglected. So fine hairs removal is crucial for improving segmentation accuracy of skin images and overall performance of skin aging analysis. In this paper, we propose a region-growing based algorithm for automatic detection of fine hairs. Skin images are first segmented coarsely with horizontal and vertical difference operator. Seeds are selected by morphological thinning operation together with length thresholding processing. Region growing algorithm is then designed to detect the fine hair pixels according to their local intensity continuity. Experimental results show that our algorithm can extract fine hairs at a reasonable speed with high accuracy and is robust to image noise.