基于特性响度的心理声学特征在车辆目标识别中的应用

Application of Specific-loundness-based Psychoacoustic Features in Automobile Recognition

为满足现代交通管理的需求,针对车辆目标识别提出了基于人耳主观反应的特性响度类特征作为识别特征,用神经网络分类器对不同类型汽车的声音信号进行的仿真研究,实现了汽车类型的自动识别。

Aimed at vehicle traffic management, a new kind of auditory features based on psychoacoustics parameters is applied to automobile recognition. Specific loudness features are calculated as recognition features in this paper. Target is recognized with neural network classifier. Results of a great number of simulations show that specific loudness features are valid for recognizing automobile types.