VeriChain: Blockchain-Based Document Verification and Authentication System Using AI

Abstract

Blockchain-based document verification systems provide secure and tamper-proof validation but have lim-itations like sensitivity to visual changes and no intelli-gent forgery detection. This paper presents VeriChain, a strong document authentication framework that combines color-invariant hashing and CNN-based forgery detection. A standard preprocessing pipeline ensures consistent hash generation despite differences in color, format, or scanning; a Convolutional Neural Network identifies tampered areas and structural problems. By using SHA-512 hashing, RSA-based signatures, and AI-driven validation, the proposed system improves accuracy, security, and reliability. Ex-perimental results show better performance compared to traditional blockchain-based verification methods, making VeriChain suitable for real-world use.