基于小波包变换的语音增强算法研究
|
摘要
语音信号处理的实际应用中,不可避免地会受到来自周围环境噪声的影响从而导致语音质量的下降。语音增强的目的就是从带噪语音中提取尽可能纯净的原始语音。
小波理论是一门新兴的时频分析技术,是分析类似于语音信号、地震信号等非平稳信号的有力工具。小波阈值去噪的主要思想是当含噪信号经小波变换由时域变换到小波域时,信号的小波系数相对集中在有限的区域内,而噪声的小波系数将分散到整个小波域。因此,即使输入信噪比比较低,信号变换后的小波系数也要大于噪声的小波系数。此时,可采用适当的阈值函数,在小波域内去除噪声系数,保留信号的系数,再由剩余的系数进行小波重构,即可恢复信号,达到去噪的目的。
本文在小波阈值去噪方法的基础上,提出了一种新的基于人耳听觉特性的小波包变换语音增强算法。首先,我们将带噪语音进行Bark尺度小波包变换,从而很好地模拟人耳的听觉特性;接着通过分析小波语音增强中传统的软、硬阈值函数的缺点并结合大量的实验仿真,我们构造了一个新的阈值函数,实验结果表明新的阈值函数的增强效果比传统的阈值函数有了较大的改善。
通过使用Matlab软件平台,我们对算法进行了实现。大量的仿真结果表明,我们提出的基于人耳听觉特性的小波包变换语音增强算法在主观和客观两方面都取得了较好的增强效果。
In many speech processing applications, it is very common to find the degradation of the quality of speech caused by undesirable background noise. The goal of speech enhancement is to recover original speech signals from noisy observations.
Wavelet theory is a newly developed time-frequency analysis technique and is especially of interest for the analysis of non-stationary signal such as speech, sonar seismic signal, etc. The main idea of wavelet thresholding lies in that when noising signal transforms from time domain to wavelet domain, the signal’s wavelet coefficients will spread to all area of wavelet domain. Although the energy of noise is bigger than the signal, its wavelet coefficients are smaller than the signal’s. So we can use thresholding function to cut off the coefficients of noise and use the rest of coefficients to reconstruct the denoising signal.
This paper presents a new algorithm for speech enhancement based on wavelet thresholding method. First, we decompose the noisy speech by the Bark-scaled Wavelet Packet (BS-WPD) to simulate the human auditory characteristics. Then we propose a new thresholding function which has many advantages over soft and hard thresholdings put forward by D.L. Donoho and I.M. Johnstone. Simulation result proves that the proposed new thresholding function has a better improvement.
At last, simulation of the algorithm based on Matlab software is implemented. A large amount of simulation results indicate that our new method based on Bark-scaled wavelet packet decomposition has a better performance in both objective and subjective aspects.
小波理论是一门新兴的时频分析技术,是分析类似于语音信号、地震信号等非平稳信号的有力工具。小波阈值去噪的主要思想是当含噪信号经小波变换由时域变换到小波域时,信号的小波系数相对集中在有限的区域内,而噪声的小波系数将分散到整个小波域。因此,即使输入信噪比比较低,信号变换后的小波系数也要大于噪声的小波系数。此时,可采用适当的阈值函数,在小波域内去除噪声系数,保留信号的系数,再由剩余的系数进行小波重构,即可恢复信号,达到去噪的目的。
本文在小波阈值去噪方法的基础上,提出了一种新的基于人耳听觉特性的小波包变换语音增强算法。首先,我们将带噪语音进行Bark尺度小波包变换,从而很好地模拟人耳的听觉特性;接着通过分析小波语音增强中传统的软、硬阈值函数的缺点并结合大量的实验仿真,我们构造了一个新的阈值函数,实验结果表明新的阈值函数的增强效果比传统的阈值函数有了较大的改善。
通过使用Matlab软件平台,我们对算法进行了实现。大量的仿真结果表明,我们提出的基于人耳听觉特性的小波包变换语音增强算法在主观和客观两方面都取得了较好的增强效果。
In many speech processing applications, it is very common to find the degradation of the quality of speech caused by undesirable background noise. The goal of speech enhancement is to recover original speech signals from noisy observations.
Wavelet theory is a newly developed time-frequency analysis technique and is especially of interest for the analysis of non-stationary signal such as speech, sonar seismic signal, etc. The main idea of wavelet thresholding lies in that when noising signal transforms from time domain to wavelet domain, the signal’s wavelet coefficients will spread to all area of wavelet domain. Although the energy of noise is bigger than the signal, its wavelet coefficients are smaller than the signal’s. So we can use thresholding function to cut off the coefficients of noise and use the rest of coefficients to reconstruct the denoising signal.
This paper presents a new algorithm for speech enhancement based on wavelet thresholding method. First, we decompose the noisy speech by the Bark-scaled Wavelet Packet (BS-WPD) to simulate the human auditory characteristics. Then we propose a new thresholding function which has many advantages over soft and hard thresholdings put forward by D.L. Donoho and I.M. Johnstone. Simulation result proves that the proposed new thresholding function has a better improvement.
At last, simulation of the algorithm based on Matlab software is implemented. A large amount of simulation results indicate that our new method based on Bark-scaled wavelet packet decomposition has a better performance in both objective and subjective aspects.
引文
[1]赵力,语音信号处理,北京:机械工业出版社,2003
[2]易克初,田斌,付强,语音信号处理,北京:国防工业出版社,2000
[3]胡航,语音信号处理,哈尔滨:哈尔滨工业大学出版社,2000
[4]杨行峻,迟惠生等,语音信号数字处理,北京:电子工业出版社,1995
[5]王晶,傅丰林,张运伟,语音增强算法综述,声学与电子工程,2005,77(1):22~26
[6]李建平,杨万年译,Ingrid Daubechies 著,小波十讲,北京:国防工业出版社,2004
[7]冯象初,甘小冰,宋国乡,数值泛函与小波理论,西安:西安电子科技大学出版社,2003
[8]成礼智,郭汉伟,小波与离散变换理论及工程实践,北京:清华大学出版社,2005
[9]徐长发,李国宽,实用小波方法,武汉:华中科技大学出版社,2004
[10]秦前清,杨宗凯,实用小波分析,西安:西安电子科技大学出版社,1994
[11]钱世锷,时频变换与小波变换导论,北京:机械工业出版社,2005
[12]江铭炎,郝宇,基于小波变换的语音增强去噪方法,山东大学学报,2001,6:201~204
[13]Jansen M , Noise reduction by wavelet thresholding, New York: Springer-Verlag, 2001
[14]D.L.Donoho and I.M Johnstone, Adapting to unknown smoothness via wavelet shrinkage, Journal of American Stat Assoc,1995,12(90),1200~1224
[15]D.L.Donoho and I.M Johnstone, Ideal spatial adaptation by wavelet shrinkage, Biometrika,Vol.81,No.3,1994,425~455
[16]王晶,傅丰林,陈建,基于小波变换多阈值语音增强处理研究, 声学与电子工程,2004,76(4):32~35
[17]孙延奎,小波分析及其应用,北京:机械工业出版社 ,2005
[18]胡广书,现代信号处理教程,北京:清华大学出版社,2004
[19]杨福生,小波变换的工程分析与应用,北京:科学出版社,2000
[20]宗孔德,多抽样率信号处理,北京:清华大学出版社,1995
[21]王炜,杨道淳,方元等,基于听觉模型的小波包变换的语音增强,南京大学学报(自然科学),2001,37(5):630~636
[22]Israel Cohen, Enhancement of speech using Bark-Scaled wavelet packet decomposition, Eurospeech-2001,September 2001,3~7
[23]Nathalie Virag, Single channel speech enhancement based on masking properties of the human auditory system, IEEE Trans on speech and audio processing,Vol.7,No.2,March 1999
[24]徐爽,韩芳芳,郑德忠,基于阈值的小波域语音增强新算法,传感技术学报,2004,1:150~153
[25]陶智,赵鹤鸣,龚呈卉,基于听觉掩蔽效应和 Bark 子波变换的语音增强,声学学报,2005,30(4):367~372
[26]D.L.Donoho, Denoising by soft thresholding, IEEE Trans. on Information Theory, Vol.41,1995
[27]I.M Johnstone, B.W Sliverman, Wavelet threshold estimators for data with correlated noise, J. Poy. Statist. Soc,B,Vol.59,1997,319~351
[28]A.Lallouani, M.Gabrea and C.S.Gargour, Wavelet based speech enhancement using two different threshold-based denoising algorithms, Canadian Conference on Electronical and Computer Engineering, Vol.1,May 2004
[29]Saeed Ayat, Mohammad T.Manzuri,Roohollah Dianat, Wavelet based speech enhancement using a new thresholding algorithm, Proceedings of 2004 International Symposium on Intelligent Multimedia, Video and Speech Processing,October 2004
[30]张维强,宋国乡,基于一种新的阈值函数的小波域信号去噪,西安电子科技大学学报(自然科学版),2004,31(2):296~303
[31] 王振力,张雄伟,郑翔,一种新的子波域语音增强方法,信号处理,2006,22(3),325~328
[32]董长虹,高志,余啸海,Matlab 小波分析工具箱原理与应用,北京:国防工业出版社,2004
[33]D.G Childers,Matlab 之语音处理与合成工具箱(影印版),北京:清华大学出版社,2005
[34]飞思科技产品研发中心,小波分析理论与 MATLAB7 实现,北京:电子工业出版社,2005
[35]张志涌,精通 Matlab6.5 版,北京:北京航空航天大学出版社,2003
[36]陈怀琛,吴大正,高西全,Matlab 在电子信息课程中的应用,北京:电子工业出版社,2002
[37]胡志华,基于 MATLAB 6.X 的系统分析与设计 小波分析,西安:西安电子科技大学出版社,2004
[38]何坤,李健,乔强等,非平稳环境下基于小波变换的信号去噪,信号处理,2005,21(3):244~248
[39]曲天书,戴逸松,王树勋,基于 SURE 无偏估计的自适应小波阈值去噪,电子学报,2002,30(2):266~268
[40] 马晓红,宋辉,殷福亮,自适应小波阈值语音增强新方法,大连理工大学学报,2006,46(4):561~566
[41] 陈立伟,赵春晖, 姜海丽,基于自适应提升小波变换的语音增强算法的研究,哈尔滨工程大学学报,2005,26(5),668~671
[42]Sungwook Chang, Y.Kwon and Sung-il Yang, Speech enhancement for non-stationary noise environment by adaptive wavelet packet, Acoustics, Speech, and Signal Processing, 2002(ICASSP '02), Vol.1,2002,561~564
[43]Yu Shao, Chip-Hong Chang, A versatile speech enhancement system based on perceptual wavelet denoising, Circuits and System, Vol.2,May 2005
[44]Ching P C, So H C, Wu S Q, On wavelet denoising and its application to time delay estimation, IEEE Trans on SP, 1999,47(10),1879~2882
[45]陶智,葛良,基于减谱法的语音增强和噪声消除的研究,苏州大学学报(自然科学),2002,18(7):58~61
[46]王让定,柴佩琪,一种基于改进减谱法的语音增强方法,模式识别与人工智能,2003,16(2):47~251
[47]楼红伟,胡光锐,基于简化的 KTL 和小波变换的非平稳宽带噪声语音增强,控制与决策,2003,18(5):577~580
[48]陶智,赵鹤鸣,龚呈卉等,基于谱减法的听觉模拟的语音增强,计算机工程与应用,2005.4:57~59
[49]Rainer Martin, Noise power spectral density estimation based on optimal smoothing and minimum statistics, IEEE Trans on speech and audio processing, Vol.9,No.5,July 2001
[50]Steven F.Boll, Suppression of acoustic noise in speech using spectral substraction, IEEE Trans on acoustics, speech, and signal processing,Vol.27,No.2,April 1979
[51]赵瑞珍,宋国乡,王红,小波系数阈值估计的改进模型,西北工业大学学报,2001,19(4):625~628
[52]Junpei Yamauchi, Tetsuya Shimamura, Noise estimation using high frequency regions for speech enhancement in low SNR environments, Speech Coding 2002, IEEE Workshop Proceedings, 6-9.Oct 2002,59~61
[53]L.Lin, W.H. Holmes and E.Ambikairajah, Adaptive noise estimation algorithm for speech enhancement, Electronics Letters, Vol.39,No.9,May 2003,754~755
[2]易克初,田斌,付强,语音信号处理,北京:国防工业出版社,2000
[3]胡航,语音信号处理,哈尔滨:哈尔滨工业大学出版社,2000
[4]杨行峻,迟惠生等,语音信号数字处理,北京:电子工业出版社,1995
[5]王晶,傅丰林,张运伟,语音增强算法综述,声学与电子工程,2005,77(1):22~26
[6]李建平,杨万年译,Ingrid Daubechies 著,小波十讲,北京:国防工业出版社,2004
[7]冯象初,甘小冰,宋国乡,数值泛函与小波理论,西安:西安电子科技大学出版社,2003
[8]成礼智,郭汉伟,小波与离散变换理论及工程实践,北京:清华大学出版社,2005
[9]徐长发,李国宽,实用小波方法,武汉:华中科技大学出版社,2004
[10]秦前清,杨宗凯,实用小波分析,西安:西安电子科技大学出版社,1994
[11]钱世锷,时频变换与小波变换导论,北京:机械工业出版社,2005
[12]江铭炎,郝宇,基于小波变换的语音增强去噪方法,山东大学学报,2001,6:201~204
[13]Jansen M , Noise reduction by wavelet thresholding, New York: Springer-Verlag, 2001
[14]D.L.Donoho and I.M Johnstone, Adapting to unknown smoothness via wavelet shrinkage, Journal of American Stat Assoc,1995,12(90),1200~1224
[15]D.L.Donoho and I.M Johnstone, Ideal spatial adaptation by wavelet shrinkage, Biometrika,Vol.81,No.3,1994,425~455
[16]王晶,傅丰林,陈建,基于小波变换多阈值语音增强处理研究, 声学与电子工程,2004,76(4):32~35
[17]孙延奎,小波分析及其应用,北京:机械工业出版社 ,2005
[18]胡广书,现代信号处理教程,北京:清华大学出版社,2004
[19]杨福生,小波变换的工程分析与应用,北京:科学出版社,2000
[20]宗孔德,多抽样率信号处理,北京:清华大学出版社,1995
[21]王炜,杨道淳,方元等,基于听觉模型的小波包变换的语音增强,南京大学学报(自然科学),2001,37(5):630~636
[22]Israel Cohen, Enhancement of speech using Bark-Scaled wavelet packet decomposition, Eurospeech-2001,September 2001,3~7
[23]Nathalie Virag, Single channel speech enhancement based on masking properties of the human auditory system, IEEE Trans on speech and audio processing,Vol.7,No.2,March 1999
[24]徐爽,韩芳芳,郑德忠,基于阈值的小波域语音增强新算法,传感技术学报,2004,1:150~153
[25]陶智,赵鹤鸣,龚呈卉,基于听觉掩蔽效应和 Bark 子波变换的语音增强,声学学报,2005,30(4):367~372
[26]D.L.Donoho, Denoising by soft thresholding, IEEE Trans. on Information Theory, Vol.41,1995
[27]I.M Johnstone, B.W Sliverman, Wavelet threshold estimators for data with correlated noise, J. Poy. Statist. Soc,B,Vol.59,1997,319~351
[28]A.Lallouani, M.Gabrea and C.S.Gargour, Wavelet based speech enhancement using two different threshold-based denoising algorithms, Canadian Conference on Electronical and Computer Engineering, Vol.1,May 2004
[29]Saeed Ayat, Mohammad T.Manzuri,Roohollah Dianat, Wavelet based speech enhancement using a new thresholding algorithm, Proceedings of 2004 International Symposium on Intelligent Multimedia, Video and Speech Processing,October 2004
[30]张维强,宋国乡,基于一种新的阈值函数的小波域信号去噪,西安电子科技大学学报(自然科学版),2004,31(2):296~303
[31] 王振力,张雄伟,郑翔,一种新的子波域语音增强方法,信号处理,2006,22(3),325~328
[32]董长虹,高志,余啸海,Matlab 小波分析工具箱原理与应用,北京:国防工业出版社,2004
[33]D.G Childers,Matlab 之语音处理与合成工具箱(影印版),北京:清华大学出版社,2005
[34]飞思科技产品研发中心,小波分析理论与 MATLAB7 实现,北京:电子工业出版社,2005
[35]张志涌,精通 Matlab6.5 版,北京:北京航空航天大学出版社,2003
[36]陈怀琛,吴大正,高西全,Matlab 在电子信息课程中的应用,北京:电子工业出版社,2002
[37]胡志华,基于 MATLAB 6.X 的系统分析与设计 小波分析,西安:西安电子科技大学出版社,2004
[38]何坤,李健,乔强等,非平稳环境下基于小波变换的信号去噪,信号处理,2005,21(3):244~248
[39]曲天书,戴逸松,王树勋,基于 SURE 无偏估计的自适应小波阈值去噪,电子学报,2002,30(2):266~268
[40] 马晓红,宋辉,殷福亮,自适应小波阈值语音增强新方法,大连理工大学学报,2006,46(4):561~566
[41] 陈立伟,赵春晖, 姜海丽,基于自适应提升小波变换的语音增强算法的研究,哈尔滨工程大学学报,2005,26(5),668~671
[42]Sungwook Chang, Y.Kwon and Sung-il Yang, Speech enhancement for non-stationary noise environment by adaptive wavelet packet, Acoustics, Speech, and Signal Processing, 2002(ICASSP '02), Vol.1,2002,561~564
[43]Yu Shao, Chip-Hong Chang, A versatile speech enhancement system based on perceptual wavelet denoising, Circuits and System, Vol.2,May 2005
[44]Ching P C, So H C, Wu S Q, On wavelet denoising and its application to time delay estimation, IEEE Trans on SP, 1999,47(10),1879~2882
[45]陶智,葛良,基于减谱法的语音增强和噪声消除的研究,苏州大学学报(自然科学),2002,18(7):58~61
[46]王让定,柴佩琪,一种基于改进减谱法的语音增强方法,模式识别与人工智能,2003,16(2):47~251
[47]楼红伟,胡光锐,基于简化的 KTL 和小波变换的非平稳宽带噪声语音增强,控制与决策,2003,18(5):577~580
[48]陶智,赵鹤鸣,龚呈卉等,基于谱减法的听觉模拟的语音增强,计算机工程与应用,2005.4:57~59
[49]Rainer Martin, Noise power spectral density estimation based on optimal smoothing and minimum statistics, IEEE Trans on speech and audio processing, Vol.9,No.5,July 2001
[50]Steven F.Boll, Suppression of acoustic noise in speech using spectral substraction, IEEE Trans on acoustics, speech, and signal processing,Vol.27,No.2,April 1979
[51]赵瑞珍,宋国乡,王红,小波系数阈值估计的改进模型,西北工业大学学报,2001,19(4):625~628
[52]Junpei Yamauchi, Tetsuya Shimamura, Noise estimation using high frequency regions for speech enhancement in low SNR environments, Speech Coding 2002, IEEE Workshop Proceedings, 6-9.Oct 2002,59~61
[53]L.Lin, W.H. Holmes and E.Ambikairajah, Adaptive noise estimation algorithm for speech enhancement, Electronics Letters, Vol.39,No.9,May 2003,754~755