Development of an Autonomous Face Recognition Machine.

Abstract : A computer algorithm was developed which successfully locates and identifies human face(s) that are present in a digitized computer image. In the process of finding the facial image, the algorithm simultaneously determines the boundary locations of the sides of the eyes, the center of the face, the tip of the nose and the vertical center of the mouth. Detection of facial images is based on analysis a digitized scene for the presence of characteristic facial feature signatures for the eyes, nose and mouth. These signatures are generated by the application of a 'center of mass' calculation to each pixel row and column for various sub-sections of the digitized scene. The presence of a face is confirmed, and its feature locations are determined based on the presence and location of the local maxima and minima which occur in these curvilinear signatures. Keywords: image processing; artificial intelligence; theses.