A Study for Reconstruction of Face recognitions through the Eigen-Algorithm

Abstract

This paper introduces the application of a new algorithm developed by the author that has solved the eigenproblem to find the eigenface, which is created in face recognition, within the process of reconstruction processing of face recognition. The reliability and limit of error of this algorithm can be observed from the face reconstruction results. Also, parallel processing can be performed through this algorithm in the given face recognition problem. A simulation was operated on 25 face images with different motions and various forms. Furthermore, this paper presents the process of comparing the original image with the reconstructed image. Moreover, it presents the advantage of creating the eigenvalues of the largest values through the algorithm of solving the eigenproblem, which the author introduces, in the process of sequentially creating the eigenvalues of the largest values, which is demanded in principle component algorithm of face recognition.