An Efficient Approach to Skin Color Segmentation Using Perceptually Uniform CIELab Color Model

The skin segmentation is one of the most crucial steps in many applications such as face detection, human identification, and video conferencing. However, it is not straightforward to classify the skin pixels because the skin tone is extensively changing from darkest to lightest due to the divergence in the quantity of melanin (pigmentation). The proposed work focused on the segmentation of skin pixels using the device independent and perceptually uniform CIELab color model. In the proposed method, all the three components of the CIELab model are independently applied for skin segmentation. The outcome of the CIELab-based skin segmentation is compared to the result of RGB-based segmentation to exhibit the preeminence of the proposed method. The investigational outcome clearly exemplified that the chrominance components ‘a’ and ‘b’ precisely classified skin pixels in all test images.