A.G. Khalin

Problem statement. Existing biometric systems (hereinafter - BS) are based on the measurement of physiological and anatomical distinguishing characteristics of people. The BS compares the presented biometric samples (photographs) and determines the degree of similarity, which is used to make a decision on admission to the object or to classified information. One of the types of attacks on biometric presentation, against which no system is fully protected, is the method of morphing. It is based on the creation of an artificial biometric sample from two original ones, which retains the characteristics of both at a level sufficient to pass the biometric algorithm with high degrees of similarity.

The purpose of the study. To develop a methodology for detecting digital morphed photographic portraits based on the artifact segmentation model for solving biometric identification problems.

Results. A methodology for detecting digital morphed photographic portraits based on an artefact segmentation model in RGB colour space is proposed.

Practical significance. The obtained methodology can be used in the BS before the biometric algorithm in order to protect the system from the corresponding type of attacks on these systems.

Khalin A.G. A methodology for detecting digitally morphed photographic portraits based on an artifact segmentation model. Radiotekhnika. 2024. V. 88. № 2. P. 25−30. DOI: https://doi.org/10.18127/j00338486-202402-04 (In Russian)

1. Nemtinov D.M., Mishhuk A.V., Antonov O.F. Metodika sravnenija biometricheskih algoritmov v programmno-tehnicheskih kompleksah. Radiotehnika. 2021. № 2. S. 64–68. DOI: 10.18127/j00338486-202102-10 (in Russian).
2. GOST 1979–5–2013 Formaty obmena biometricheskimi dannymi. Ch. 5. Dannye izobrazhenija lica. M.: Standartinform. 2015. 109 s. (in Russian)
3. Dvojris L.I., Krjukov I.N. Neparametricheskij metod ocenki odnomernyh plotnostej raspredelenija verojatnosti jeksperimental'nyh dannyh. Radiotehnika. 2022. № 1. S. 11–15. DOI: https://doi.org/10.18127/j00338486-202201-02 (in Russian).
4. Viola P., Jones M. Rapid Object Detection using a Boosted Cascade of Simple Features. 2nd International Workshop on Statistical and Computational Theories of Vision-Modelling, Learning, Computing, and Sampling. 2001.
5. Gonsales P., Vuds R. Mir cifrovoj obrabotka izobrazhenij. M.: Tehnosfera. 2019. 1103 s. (in Russian).