基于Frost与MFCC特征的车型识别方法研究

Research on Vehicle Recognition Method Based on Frost and MFCC Feature

车辆识别在智能交通系统中占有重要地位,在交通数据统计、自动收费系统等相关领域得到广泛应用。为克服图像识别方法的局限性,提出了一种将麦克风阵列声音增强技术与声音识别技术相结合的新型车辆识别技术。首先利用麦克风阵列的空间滤波特性,将目标信号与干扰信号分离;其次对信号进行MFCC特征提取;最后通过分类器识别声信号类型。实验结果表明,Frost波束形成器能有效增强目标信号,抑制干扰信号,具有良好的指向性,可使卡车识别准确率达到93.1%。该新型车辆识别方法能够有效进行分类识别,对该方向的研究具有实用价值。

Vehicle recognition plays an important role in intelligent transportation systems and is widely used in related fields such as traffic data statistics and automatic toll collection systems.To overcome the limitations of image recognition methods,a new vehicle recognition technology that combines microphone array sound enhancement technology and sound recognition technologyis proposed.Firstly,the spatial filtering characteristics of the microphone array are utilized to separate the target signal from the interference signal.Secondly,MFCC feature extraction is performed from the signal.Finally,the acoustic signal type is identified through a classifier.The experimental results show that the Frost beamformer can effectively enhance the target signal,suppress interference signals,and have good directionality,achieving a recognition accuracy of 93.1%for trucks.This new vehicle recognition method can effectively perform classification recognition and has practical value for research in this direction.