Blockchain-Enabled CCTV Integrity Framework with Anomaly Detection for Microfinance Banks in Nigeria

Abstract

Closed-Circuit Television (CCTV) systems are widely used in Nigerian financial institutions to enhance security, monitor transactions, and ensure regulatory compliance. Conventional centralized CCTV architectures, however, are vulnerable to tampering, insider threats, and single points of failure, undermining the reliability of video evidence. This paper presents a blockchain-enabled CCTV integrity framework integrating machine learning-based anomaly detection, specifically designed for resource-constrained microfinance banks in Nigeria. Cryptographic hashes of CCTV footage are anchored on a permissioned Ethereum blockchain to ensure immutability and chain-of-custody, while full video content is stored off-chain in a SQL or IPFS repository. A Flask-based dashboard facilitates secure upload, verification, and retrieval of CCTV files, with automatic tamper alerts generated by the anomaly detection module. A prototype was implemented using Python, Web3.py, Ganache (PoA), and MySQL and evaluated with 30 video samples (15 original, 15 tampered). Performance metrics included hash generation time, blockchain write latency, verification accuracy, anomaly detection efficiency, and system resilience under constrained network and power conditions. Results indicate that the proposed framework improves tamper resistance, evidentiary integrity, and operational reliability compared to centralized and hash-only schemes, while remaining cost-effective. The study demonstrates that integrating blockchain with anomaly detection provides a practical, scalable, and secure solution for CCTV surveillance in Nigerian microfinance banks.