基于GMM和枪声的军事环境判别

枪声在军事环境下是一个令人高度重视的声音信号。针对其所含信息量的重要性，提出了一种基于枪声检测的军事环境判断方法。首先利用基于中值滤波的算法检测出枪声的起止点；采取反应人耳听觉机理的声音信号动态特征-Mel倒谱系数描述声音信号；然后利用基于高斯混合模型的枪声检测及识别系统完成对枪声的分析从而实现对当前军事环境的判断。实验结果证明了该方法的准确性和有效性。

Military Environment Discrimination Based on GMM and Gunshot Detection

The sound of gunshots is very important for the military environment.According to the importance of gunshots,a novel method is proposed to discriminate the military environment,based on the detection of gunshots.First,the start and the end of gunshots is detected utilizing the median filtering.And then the MFCC parameter is choosed to describe the sound signal.The military environment is discriminated by the recognition of gunshots with the help of GMM.The experimental results conducted on the database of g...