20.500.12556/RUL-177797 Biometrično prepoznavanje posameznih živali znotraj vrst družine Felidae Biometric recognition of individual animals within species from the Felidae family Področje slikovne biometrije se ukvarja s prepoznavo posameznikov glede na njihove unikatne biološke lastnostni z uporabo slik. Večina raziskav temelji na prepoznavi ljudi, manj pa se jih osredotoča na preostanek živalskega kraljestva, predvsem na divje vrste. Želeli smo raziskati, kako dobro prepoznava deluje nad eno izmed teh živalskih družin. V magistrski nalogi raziskujemo možnosti uporabe slikovne biometrije in globokih nevronskih mrež, specifično siamskih mrež za prepoznavo predstavnikov živalske skupine divjih mačk Felidae. Zgradimo več naborov podatkov za potrebe treniranja modelov, ki vključujejo različne modalnosti – celoten očesni predel, šarenica, vzorci dlake in vzorci dlake za posamezne dele telesa. Vse nabore treniramo z več modeli in hiperparametri z uporabo kontrastne izgube ter opazujemo točnost parov slik, ki so napovedani kot pravilno ujemanje. Najbolje deluje prepoznava nad celotnim očesnim predelom, nato sledi šarenica in nazadnje vzorci dlake. Vsi omenjeni nabori podatkov dosežejo točnost nad 85%, najbolje odrezani očesni predel pa doseže 92.76% točnost z uporabo ResNet18 modela in brez bogatenja podatkov. The field of image-based biometry focuses on recognition of individuals based on their unique biological characteristics with the usage of images. Most research is based on human recognition, while less work is focused on the remaining animal kingdom, especially wild species. We wanted to explore how well recognition works on one of these animal families. In the master’s thesis, we explore the possibilities of using image-based biometrics and deep neural networks, specifically Siamese networks, for the recognition of members of the wild cat family Felidae. We construct several datasets for teaching the models, which include different biometric modalities – the entire eye region, the iris, fur patterns and fur patterns of specific body parts. All datasets are trained with multiple models and hyperparameters using contrastive loss. We evaluate the accuracy of image pairs predicted as genuine matches. The best performance is achieved with recognition based on the entire eye region, followed by the iris, and finally fur patterns. All datasets reach accuracy above 85%, with the eye region performing best, achieving 92.76% accuracy using the ResNet18 model without data augmentation. felidae slikovna biometrija nevronske mreže globoke nevronske mreže siamske nevronske mreže konvolucijske nevronske mreže felidae image-based biometry neural networks deep neural networks siamese neural networks convolution neural networks true false false Slovenski jezik Angleški jezik Magistrsko delo/naloga 2026-01-07 15:51:17 2026-01-07 15:51:17 2026-07-02 08:44:21 0000-00-00 00:00:00 2025 0 0 0000-00-00 NiDoloceno NiDoloceno NiDoloceno 0000-00-00 0000-00-00 0000-00-00 37747 283330819 Matek_Veronika_-_Biometricno_prepoznavanje_posameznih_zivali_znotraj_vrst_druzine_Felidae.pdf Matek_Veronika_-_Biometricno_prepoznavanje_posameznih_zivali_znotraj_vrst_druzine_Felidae.pdf 1 AAE9F0FB94E146ACF0A62811597969AD d102138aef6bfb76452b0372b6a970f31baed93f8f4bb9157a3eb01163d6f455 51313718-ebd8-11f0-bf92-0050569b8976 https://repozitorij.uni-lj.si/Dokument.php?lang=slv&id=225124 Fakulteta za računalništvo in informatiko 0 0 0