Homomorphic Encryption Based Video Copy Detection In Multi View Videos Stored In A Cloud
DOI:
https://doi.org/10.62643/Keywords:
Homomorphic Encryption, Video Copy Detection, Multi-View Video, Cloud Computing, Privacy- Preserving Computation, Content-Based Video Retrieval, Secure Video Processing, Encrypted Domain Processing, Digital Media Forensics, Cloud Storage SecurityAbstract
The rapid proliferation of multimedia content and the increasing adoption of cloud storage have led to
growing concerns over unauthorized duplication and distribution of video content. This issue becomes
particularly complex with multi-view videos, which offer multiple perspectives of the same scene and are
frequently used in surveillance, entertainment, and immersive experiences. Ensuring the integrity and
ownership of such content requires robust and privacy-preserving mechanisms. This project proposes a novel
framework for detecting video copies in a multi-view environment while maintaining the confidentiality of
the data using homomorphic encryption. Homomorphic encryption allows computations to be performed
directly on encrypted data without the need to decrypt it, thus preserving the privacy of the content even
during analysis. This unique capability makes it an ideal candidate for video copy detection in cloud
environments, where data privacy is a critical concern. In this system, each multi-view video is processed to
generate compact and distinctive feature vectors, which are then encrypted using a homomorphic encryption
scheme. These encrypted features are stored securely in the cloud and are later used for copy detection
queries.To address the complexities of multi-view video analysis, the framework employs view-invariant
feature extraction techniques that can robustly capture the spatio-temporal characteristics of video content
across different perspectives. These features are resilient to transformations, such as rotation and scaling,
ensuring effective matching even when copies differ in visual representation. By utilizing encrypted-domain
matching algorithms, the system can compare newly uploaded or queried videos against the encrypted
database to detect potential duplicates or pirated versions.A major contribution of this work lies in the
integration of privacy-preserving cryptographic techniques with advanced multimedia content analysis in a
scalable cloud infrastructure. This ensures that sensitive video data is never exposed in plaintext form,
addressing both data security and compliance requirements. Furthermore, the system’s architecture is
designed to support real-time querying, making it practical for large-scale deployment across content
platforms and video streaming services
Downloads
Published
Issue
Section
License

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.













