An access control system based on multimodal approach

The article presents a description of an access control and management system developed on the basis of a multimodal approach to improve the accuracy of the system’s functioning by developing personality identification models. The multimodal approach is based on the simultaneous application of the developed analytical and procedural models of personal identification by fingerprints and the iris of the user’s eye. The analytical model for constructing the contour of the iris of the user’s eye is distinguished by the use of the Canny algorithm to determine the boundaries of the pupil, the median filter to eliminate small noise, the wavelet transformation to determine the boundaries and parameters of the iris, and the Laplace operator and the Gaussian filter to find key points when analysing the image. The analytical model of the neuro-fuzzy classification of the user’s signature is distinguished by the use of the Gaussian function as a radial function. A procedural model of classification, recognition and integer identity identification based on the analysis of the user’s iris is distinguished by the use of the Gabor filter to search for and identify special points in the image. A procedural model of classification, recognition and integer identity identification based on a dynamic user signature is distinguished by the use of a sigmoidal function as a function of activating neurons of a radial neural network.