Models and Methods for Matching Images in the Problem of Face Recognition

A subject area is analyzed and the topicality of the problem of face recognition is shown. Methods for matching images by applying position lines, convolutions and invariants to the group of affine transformation for 2D and 3D images are examined. Proper comparison is a necessary condition for solving the recognition problem. Proper use of the position-line method for reducing face images to the normalized form is shown. © 2018, Allerton Press, Inc