Skripsi
PROTOTIPE SISTEM KEAMANAN AKSES MASUK RUANGAN BERBASIS FINGERPRINT RECOGNITION MENGGUNAKAN ALGORITMA DEEP LEARNING
The password is the oldest security system that is still used today. However, the security system is vulnerable to cyber attacks, especially brute force. The biometric security system is one of the solutions to improve the current security system. Fingerprints are one of the most frequently used biometrics today. In this study, the fingerprint recognition system is implemented in the form of a prototype with recognition using deep learning architecture from convolutional neural network (CNN). The training results show that ResNet50 does not experience significant overfitting compared to the other two architectures with loss value 0,9935% and accuracy 100%. Offline testing uses test data on ResNet50 with variable "open for the right person" getting a value of 95.58% while EfficientNetV2M and VGG16 get a value of 93.13% and 86.76%. So that the prototype used ResNet50 as a fingerprint recognition system. However, online testing on ResNet50 still not able to generalize new data especially in the condition of damaged fingerprints.
Inventory Code | Barcode | Call Number | Location | Status |
---|---|---|---|---|
2307005495 | T119492 | T1194922023 | Central Library (Referens) | Available but not for loan - Not for Loan |
No other version available