一种基于能零积的改进端点检测算法  

AN IMPROVED ENDPOINT DETECTION ALGORITHMBASED ON ENERGY-ZERO-PRODUCT

作　　者：韦金泉 罗丽燕 王玫 刘争红[1,2] 何锐彬 莫清麟 Wei Jinquan;Luo Liyan;Wang Mei;Liu Zhenghong;He Ruibin;Mo Qinglin(Provincial Ministry of Education Key Laboratory of Cognitive Radio and Signal Processing,Guilin University of Electronic Technology,Guilin 541004,Guangxi,China;Guangxi Key Laboratory of Wireless Broadband Communication and Signal Processing,Guilin University of Electronic Technology,Guilin 541004,Guangxi,China;College of Information Science and Engineering,Guilin University of Technology,Guilin 541007,Guangxi,China)

机构地区：[1]桂林电子科技大学认知无线电与信息处理教育部重点实验室,广西桂林541004 [2]桂林电子科技大学广西无线宽带通信与信号处理重点实验室,广西桂林541004 [3]桂林理工大学信息科学与工程学院,广西桂林541007

出　　处：《计算机应用与软件》2024年第7期215-221,共7页Computer Applications and Software

基　　金：国家自然科学基金项目(61771151);广西重点研发计划项目(桂科AB17292058);广西科技基地和人才专项项目(桂科AD20159018);广西自然科学基金面上项目(2020GXNSFAA159004);桂林电子科技大学研究生教育创新计划项目(2021YCXS024)。

摘　　要：针对基于短时能量和短时过零率的传统端点检测算法在信噪比低于10 dB时其准确性明显下降的问题,提出一种基于声谱图特征增强的能零积端点检测改进算法。首先,对音频信号进行谱减法降噪,提升信号的信噪比;其次,依次对声谱图进行腐蚀、二值化和膨胀处理,以实现声谱图特征增强;最后,提取能零积特征,并利用双阈值端点检测算法对音频信号进行端点检测。实验结果表明,该算法在不同信噪比条件下可以有效捕获有环境异常音片段,具有良好的鲁棒性。To tackle the problem of the significantly reduced accuracy of the traditional endpoint detection algorithm based on short-term energy and short-term zero-crossing rate when the signal-to-noise ratio is less than 10 dB,an improved energy-zero-product endpoint detection algorithm based on the enhancement of the Sonogram feature is proposed.The spectral subtraction method was utilized to reduce the noise of the audio signal and thus improving the signal-to-noise ratio.Perform erosion,binarization and dilation were processed on the sonogram to achieve the enhancement of the spectrogram characteristics.The energy-zero-product feature was extracted and the dual-threshold endpoint detection algorithm was used to conduct endpoint detection on the audio signal.Experimental results show that the algorithm can effectively capture the fragments with ambient abnormal sound under different signal-to-noise ratio conditions,and has good robustness.

关 键 词：端点检测 能零积 低信噪比 声谱图 特征增强 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术]