An extracted human face algorithm for facial features potential distribution

A valid algorithm for extracted human faces and facial features has been put forward. This algorithm has combined the gradient mapping of binary edge image and the potential distribution of people´s facial characteristics. First of all, using the gradient mapping of horizontal direction and vertical direction of binary edge image, to determine the location of the human face and the rough location of the human eye, then using the approximation method to location the human eye´s accurate positioning; To determine the location of mouth by the gradient mapping in vertical direction of the image; Combination of the human eye position, mouth position, as well as the potential distribution of face characteristics to determine the location of the nose. At the same time, in order to verify the effectiveness of the algorithm, a series of different people in different postures has been used to validate the algorithm.