摘要
在音频检测技术研究的基础上,设计了交通事故自动报警系统,此系统根据车辆的声强和声谱等特性,通过多特征碰撞声检测算法检测是否发生了碰撞型交通事故。文章对多特征碰撞声检测算法以及碰撞声信号时域的能量分布特征进行了分析,验证了多特征碰撞声检测算法的高精度和可靠性。该研究可为实现交通事故的快速救援、降低事故中因抢救延迟造成的死亡率等提供技术支持。
Based on the research into audio detection technology, the authors of this paper have designed an automatic alarm system for traffic accidents, which detects whether a tra f -fic accident happens through the analysis of the intensity and spectrum of the sound pro-duced by a vehicle in collision and with the multi-feature collision detection algorithm. In this paper, the multi-feature collision detection algorithm and the characteristics of the time domain energy distribution of the colliding acoustic signal are analyzed and the accuracy and reliability of the multi-feature collision detection method are verified. The study can serve as technical support for the rapid rescue in a traffic accident and the reduction of the mortality caused by delayed rescue during an accident.
出处
《唐山学院学报》
2017年第3期50-54,共5页
Journal of Tangshan University
基金
国家科技支撑计划项目(2011BAK20B00)
关键词
交通事故
音频检测
自动报警
特征算法
traffic accident audio detection automatic alarm feature algorithm
