The Face Recognition Method Based On Convex Hull Algorithm

With the continuous progress of human society, there is growing emphasis on self-awareness and protection, there are also differences in the relative stability of the natural characteristics between individuals and extracted biometric naturally, being a human life is a important part of face recognition.feature extraction is one of the most biologically important aspect. Face recognition technology advances is not only human self-awareness increasing visual expression, but also great to meet the need for social security and artificial intelligence. With the emergence of innovative computer and network technology, artificial intelligence began to give people work and life,and brought a lot of convenience, as the more popular face recognition technology has been widely applied to all walks of life, this paper based on the convex hull improved algorithms for face recognition program.This paper introduces the current popular three classical face recognition matching algorithm, the algorithm though is widely used, but there are still some shortcomings and problems to be solved. This paper presents a combination of innovative applications of convex hull algorithms on face recognition.The main findings are as follows:1. classic convex hull algorithms and applications deployed around the convex hull algorithm has been studied since the convex hull algorithm is usually large areas of application and the amount of data, the lack of time on the efficiency of traditional methods exist to impede the algorithm application. After systematic research, we propose an algorithm based on the convex hull improved "bi-directional rotation screen dot stuck algorithm," the algorithm through argumentation theory and experimental data verification. Compared with the traditional convex hull algorithm in experimental screening threshold range set point, not only has a high accuracy rate, and has been greatly improved time efficiency. Finally, we give the results of the experiment.2. The proposed improvements will get "stuck screening point bi-directional rotation algorithm" which has been used in the two-dimensional face recognition method, and gave specific steps utilization. Specific application needs by artificially defined threshold calculation "on the heel point", after gradation processing of the image forming face data and combined with improved screening convex hull algorithm, resulting to improve new data, and finally compering the face data match to identify.3. Further improvements proposed of algorithm is applied to three-dimensional face recognition technology, this paper studied the method of collection grates projection 3D face images, as well as the difficulties and problems in two-dimensional three-dimensional face recognition and face recognition compared to the existence of proposed after two dimensional 3D face, using the improved algorithm proposed further experiments, the same principle of a two-dimensional recognition. The main purpose of this algorithm is proposed to solve the three-dimensional face recognition time problem of the poor efficiency.