A Novel Video Forgery Detection Model Based on Triangular Polarity Feature Classification

Chee Cheun Huang, Chien Eao Lee, Vrizlynn L. L. Thing

Source Title: International Journal of Digital Crime and Forensics (IJDCF)12(1)

ISSN: 1941-6210|EISSN: 1941-6229|EISBN13: 9781799805793|DOI: 10.4018/IJDCF.2020010102

MLA

Huang, Chee Cheun, et al. "A Novel Video Forgery Detection Model Based on Triangular Polarity Feature Classification." IJDCF vol.12, no.1 2020: pp.14-34. http://doi.org/10.4018/IJDCF.2020010102

APA

Huang, C. C., Lee, C. E., & Thing, V. L. (2020). A Novel Video Forgery Detection Model Based on Triangular Polarity Feature Classification. International Journal of Digital Crime and Forensics (IJDCF), 12(1), 14-34. http://doi.org/10.4018/IJDCF.2020010102

Chicago

Huang, Chee Cheun, Chien Eao Lee, and Vrizlynn L. L. Thing. "A Novel Video Forgery Detection Model Based on Triangular Polarity Feature Classification," International Journal of Digital Crime and Forensics (IJDCF) 12, no.1: 14-34. http://doi.org/10.4018/IJDCF.2020010102

Export Reference

Favorite Full-Issue Download

View Full Text HTML

View Full Text PDF

Abstract

Video forgery has been increasing over the years due to the wide accessibility of sophisticated video editing software. A highly accurate and automated video forgery detection system will therefore be vitally important in ensuring the authenticity of forensic video evidences. This article proposes a novel Triangular Polarity Feature Classification (TPFC) video forgery detection framework for video frame insertion and deletion forgeries. The TPFC framework has high precision and recall rates with a simple and threshold-less algorithm designed for real-world applications. System robustness evaluations based on cross validation and different database recording conditions were also performed and validated. Evaluation on the performance of the TPFC framework demonstrated the efficacy of the proposed framework by achieving a recall rate of up to 98.26% and precision rate of up to 95.76%, as well as high localization accuracy on detected forged videos. The TPFC framework is further demonstrated to be capable of outperforming other modern video forgery detection techniques available today.