基于车辆音频谱相关的车速估算技术研究

Study of vehicle speed estimation based on vehicle's audio frequency spectrum correlation

目前,大多车检技术和车速检测技术都存在着不足,如成本高、安装繁琐、不便维修和对临时调查项目适应性差等。针对这些不足,探索了一种新的车辆速度测量方法,以运动车辆的音频的多普勒效应为出发点,以音频频谱的拉伸(压缩)与多普勒效应参数——速度的关系为研究内容,用频谱相关性去评价拉伸(压缩)后的频谱相似程度,从而推算出车速。

Nowadays, there are deficiencies in most of vehicle detection and vehicle speed detection, such as high cost, complicated installation, inconvenience of maintaining and inadaptability for temporary investigating projects. To solve these deficiencies, this paper puts forward a new method of vehicle speed detection. Based on Doppler effect of moving vehicle＇ s audio, this paper utilizes the audio spectrum correlation to appraise the spectrum＇s similitude after stretch（or compressing） and then estimates the vehicle＇s speed by investigating the relationship between the audio spectrum＇s stretch（or compressing） and the Doppler effect parameternamely speed.