基于摄像头检测的防盗拍系统开发和算法研究

Camera detection based anti-piracy system development and algorithm design

针对电影院中的非法盗拍现象,提出了对智能摄像头检测系统的改造和算法设计。硬件部分的人工光源模块提供定向光照,与之相配合的软件算法部分融合层级特征检测、直方图比较、主成分分析、K最近邻分析等多种方法,对摄像头镜头反射高光模式进行检测和识别。通过实验数据集的验证,系统对摄像头敏感度高,检测准确率高,具有理论研究价值和实际应用价值。系统硬件模块和算法模块可拓展且可移植于其他应用领域。

Aiming to tackle the illegal film-recording in cinema, the hardware reconfiguration and algorithm design of an intelligent camera-detection system is introduced. The artificial light module provides directed illumination, which enables the algorithm to detect and recognize the specularity pattern on the camera lens by fusing multiple detection algorithms, e.g. cascade feature detection, histogram comparison, principal component analysis, and K-nearest neighbor method. The dataset validated the performance of the system, and illustrated its sensitivity to cameras, high accuracy, and its theoretical and practical values. The functionality of the hardware and software system can be further expanded and transplanted to other applications.