ARCHIVES

Online banking systems have become an essential part of modern financial services, allowing users to perform transactions conveniently through digital platforms. However, the increasing number of cyber threats and identity theft incidents has raised serious concerns regarding the security of online banking authentication mechanisms. Traditional methods such as passwords and PINs alone are often insufficient to prevent unauthorized access. To address these challenges, this project proposes a secure Online Banking Transaction System using vision Transformer (ViT) based Face Recognition, DeepFace-based Liveness Detection, and SHA-256 Hashing to strengthen authentication and transaction security. In the proposed system, users register their facial data and personal credentials in the banking platform. During transaction initiation, the system captures the user’s facial image through a camera and performs face recognition using the Vision Transformer (ViT) algorithm. The ViT model extracts deep facial features and generates unique facial embeddings that are compared with stored facial representations in the database to verify the user’s identity accurately. To prevent spoofing attacks such as the use of photographs, videos, or masks, the system integrates DeepFace-based liveness detection, which analyzes facial movements and real-time characteristics to confirm that the captured face belongs to a live person. Once the biometric verification is successfully completed, the user is allowed to proceed with the banking transaction. For additional security, transaction passwords and PINs are protected using the SHA-256 hashing algorithm, ensuring credentials are securely stored and verified without revealing the original values. The system performs validation checks before processing transactions to ensure secure operations. This approach improves the security and reliability of online banking while preventing fraud and unauthorized access.