The integration of object detection and classification technologies with surveillance cameras in a library setting aims to enhance security, resource management, and user experience. To address concerns related to data security and privacy, we incorporate blockchain technology, which provides a decentralized and immutable ledger for recording surveillance data. This network leverages federated learning to enable collaborative model training while preserving data privacy, enhancing the accuracy and robustness of the object detection and classification system across different environments. Our approach demonstrates the potential of combining computer vision, blockchain, and collaborative networks to create a secure, efficient, and privacy-preserving surveillance system for libraries.