The press service of MV Lomonosov Moscow State University reports that scientists from the university's Faculty of Computational Mathematics and Cybernetics have developed a method of identifying a person by iris, which works even in the presence of interference in the form of eyelashes.
The system uses neural network algorithms to build and compare key points of the iris. Image-based testing shows high recognition accuracy and resistance to eye occlusion, including eyelids and eyelashes.
The new method can be used in biometric systems for automatic identification of individuals. This method combines classical mathematical algorithms and neural network technology, which increases the reliability and accuracy of image analysis.
The press agency notes that this solution opens up new opportunities to improve biometric services.
