Estimating the Position, Number and Length of Forehead Wrinkles Using Neural Network

A novel approach towards 2D Facial Aging (FA) techniques using Neural Network (NN) is proposed in this study. This approach is proposed to automatically predicate the position, number and length of wrinkles on the forehead area. The method is divided into three main stages; the first stage is the preprocessing stage, where the forehead area is manually cropped then filtered to gain sharpness. After that, the wrinkles segmentation process is carried out using row-by-row threshold, morphological erosion and connected component labeling to accurately extract the wrinkles form the image. Finally, the NN is used to convert the detected wrinkle lengths from 2D to 3D curvature shape. The proposed method is objectively compared with other techniques and can accurately predict the position, number and length of the forehead wrinkles in different ages.