监控视频帧间篡改检测系统设计与性能优化研究

Design and Performanc Optimization of Surveillance Video Inter-frame Forgery Detection System

监控视频作为司法取证的重要依据,其真实性难以得到保证。帧间篡改作为最常见的视频篡改方式,其检测算法研究已取得大量卓越成果,但在实际场景应用中却面临着许多困难。分析司法场景下对监控视频内容真实性鉴证的现实需求,设计实现了功能完备的帧间篡改检测系统。为了解决系统实现算法的大计算量问题,研究GPU的硬件结构、CUDA的线程模型和内存模型,提出了通过CUDA使用GPU进行性能优化的方法,有效提升了系统性能。

As a significant evidence of judicial forensics, surveillance videos’ authenticity is dificult to guarantee. Inter-frame forgery is one of the most common ways for video tampering, and research on its detection algorithm now acquires a lot of excellent results. However, the application of these algorithms in the actual scene still faces lots of difficulties. Based on analysis of the actual demand for authenticity verification of surveillance video content in judicial scene, a fully functional inter-frame tamper detection system is designed and implemented. In order to solve large computational complexity problem of the algorithm in the system, the hardware structure of GPU, the thread model and memory model of CUDA are explored. The method with GPU to optimize the performance through CUDA is proposed, and via this method the system performance is effectively improved.