A Chunk-based Copy Detection Approach for Multimedia Documents

摘要

Copy detection is important to both intellectual property protection and information retrieval. Prior researches on copy detection (e.g., hash-based fingerprinting algorithm, etc.) concentrated on document-level copy detection. These researches have difficulty to detect complicated multimedia documents. In this study, a chunk-based copy detection approach is discussed. And a Fingerprint-based Heuristic Algorithm (FHA) is proposed based on fixed-length chunks and overlap of chunks. With experiments and results, the proposed approach can deal with both total copy detection and partial copy detection of multimedia documents.