录音设备识别司法鉴定技术研究  

Research on Methods of Audio Recorder Forensic Identification

作　　者：曾锦华[1] 施少培[1] 杨旭[1] 李岩[1] 卢启萌[1] 

机构地区：[1]司法部司法鉴定科学技术研究所,上海200063

出　　处：《中国司法鉴定》2014年第6期22-25,共4页Chinese Journal of Forensic Sciences

基　　金：国家"十二五"科技支撑项目(2012BAK16B06);中央级科研院所公益研究基金(GY2014G-2;GY2013G-3;GY2014Z-3)

摘　　要：目标随着手机和录音笔等数字录音设备的普及,数字录音已基本取代传统的模拟录音,成为录音司法鉴定主导性检材类型。数字录音资料作为视听资料的重要组成部分,其真实性司法鉴定新技术新方法的研究具有重要的理论意义和实际应用价值。方法研究基于录音设备识别的数字录音真实性鉴定技术,通过数字录音背景噪声片段的提取,计算录音设备相关的关键统计特征,包括采样直方图分布特征和平均频谱统计特征,并使用机器学习和模式分类方法对数字录音的载体即录音设备进行准确分类。结果实验中最高的分类准确性达到97.09%。在录音设备可分性研究成果基础上,提出应用于数字录音设备司法鉴定的可行实施方案。结论研究结果表明了基于信号统计特征分析的录音设备识别方法的可行性和准确性。Objective With the popularity of cell phones and digital recorders, traditional analog recordings have been largely replaced by digital ones in the audio forensic examination. Digital recordings are important parts of audio-visual materials, and the research on novel forensic authentication methods is of great importance both in theoretical significance and practical value. In this paper, the method of audio forensic authentication through identifying audio recorders was studied. Method Silent segments were firstly extracted in digital recordings, and the key equipment-related statistical features, i.e., sampling histogram distribution and spectral mean, were computed. Digital recordings were classified with the usage of the method in the fields of machine learning and pattern classification. Results The best classification accuracy in the experiments was up to 97.09%. On the basis of the result that digital recorders could be successfully identified, a feasible implementing scheme for forensic identification of digital recorders was proposed. Conclusion The audio recorder forensic identification based on signal statistical analysis is feasible and accurate.

关 键 词：录音设备 录音真实性鉴定 统计特征 模式分类 

分 类 号：DF795.2[医药卫生—法医学]