Tools for studying of facial identification and authentication

Abstract

One of the most common and popular among biometric techniques is biometry of face. Currently, a large number of information systems have been created using these techniques. Thus, there is a problem of studying the methods of machine face recognition during the training of relevant specialists. The aim of the work is to develop a software tools for studying the mechanism of biometric face authentication. Face is static biometry methodic. For implementation in the developed tools the method of comparison on key points was chosen. The main stages of comparing faces are: finding a face in a photo, formation of key points, normalization, formation of segments, formation of relations, comparison of ratios. A high-level cross-platform programming language C++ with the use of additional libraries was chosen to develop the software of the tools. An open face database was used to work with the tools. The tools have a demonstration nature. During its use, you can explore and study all the components of the process of biometric identifi-cation and authentication of the user by face: demonstration of the process of finding of key points on the face, forming of distances between points, calculating of their ratios and their comparison; comparison of etalon and candidate photos during the procedure of identification and authentication and response. Examples of the main processing stages of the complex and the corresponding screen forms (windows) are given. The com-position of the tools and software are developed in the work, the main examples and stages of its work are given. It is shown that the tools can be used in the educational process, for example in laboratory or practical classes.