摘要
车辆音频信号具有复杂的特征,单一特征提取方法不能全面反应该特点。为了使提取的音频信号特征能更好地反映车辆原始信号,提出了将已有的短时能量(energy)、短时傅里叶变换(STFT)及梅尔倒谱系数(MFCC)特征提取方法进行融合的方法,依据支持向量机(SVM)的分类识别算法,达到车辆识别的目的。实验表明,提出的组合方法优于单一提取方法,实现了提高识别率的目标;其中,ENERGY+MFCC组合方式效果最好。
The audio signal of vehicle has complex characteristics,and the single feature extraction method can t completely reverse the characteristics.In order to make the extracted audio signal characteristics better reflect the original signal of vehicles,a fusion method of the short-time energy and short-time Fourier transform(STFT)and Mel cepstrum coefficient(MFCC)is put forward.It achieves the goal of vehicle identification based on support vector machine(SVM)classification algorithm.The experiment results show that the proposed combination methods are superior to the single extraction method and achieve the target of raising the recognition rate.The method of ENERGY+MFCC is the best combination of the other ones.
作者
赵宏旭
杨文帅
ZHAO Hong-xu;YANG Wen-shuai(College of Information Engineering,Engineering University of PAP,Xi an 710086,China)
出处
《科学技术与工程》
北大核心
2018年第18期197-201,共5页
Science Technology and Engineering