BlockImage: A Secure Framework for Image Authentication and Provenance using AI and Blockchain

In contemporary applications, especially digital forensics, intellectual property protection, and secure image sharing, it is essential to guarantee the security, integrity, and authenticity of digital images. To improve image authentication, this research presents BlockImage, a sophisticated architecture that combines blockchain storage, cryptographic hashing, AI-driven information extraction, and decentralized image retrieval through IPFS. After being refined on the modified CASIA Tampered Image Dataset, a ResNet-50 model outperformed traditional techniques with a tamper detection accuracy of 94.7%. The solution maintains immutable provenance tracking using the Hyperledger Fabric blockchain and effectively identifies modifications using SHA-256 cryptographic hashing. Furthermore, tamper-proof access to images is made possible through decentralised storage through IPFS, guaranteeing an average retrieval time of about 200 ms per image. Comparing experimental assessments to current methods reveals improved security, storage efficiency, and verification capabilities. The BlockImage framework offers a high-performance, scalable way to safeguard digital images from unwanted changes, guaranteeing their reliability and accessibility over time.