Dissertation > Excellent graduate degree dissertation topics show

The Research on Speech Enhancement Based on Wavelet Packet Transform Algorithms

Author: WangWenLiang
Tutor: GuoJiChang
School: Tianjin University
Course: Signal and Information Processing
Keywords: Speech Enhancement Wavelet packet transform Human auditory characteristics Threshold function
CLC: TN912.35
Type: Master's thesis
Year: 2007
Downloads: 180
Quote: 4
Read: Download Dissertation

Abstract


In the practical application of the speech signal processing , will inevitably be affected by the influence from the surrounding environment noise resulting in a decline in voice quality . The purpose of speech enhancement is to be extracted from the noisy speech 's original voice as pure as possible . Wavelet theory is a new time-frequency analysis , a powerful tool for the analysis of non-stationary signals similar to the speech signal , the seismic signals . The main idea of wavelet threshold denoising noisy signal by wavelet transform from the time domain to the wavelet domain , the wavelet coefficients of the signal is relatively concentrated in a limited area , and noise wavelet coefficients will be dispersed throughout the wavelet domain . Therefore, even when the input SNR is relatively low , the wavelet coefficients in the signal after conversion should be greater than the wavelet coefficients of the noise . In this case, the appropriate threshold function can be used , the purpose the noise removal wavelet domain coefficients to retain the coefficient of the signal , and then by the remaining coefficient wavelet reconstruction to restore the signal , to achieve denoising . In this paper, wavelet threshold denoising method based on a new wavelet packet transform speech enhancement algorithm based on human auditory characteristics . First, we will take the noisy speech Bark scale wavelet packet transform to simulate human hearing characteristics ; followed by analysis Wavelet voice enhancement in traditional soft and hard threshold function faults and combined with a large number of experimental simulation , we construct a new threshold function , the experimental results show that the new threshold function enhancement has improved considerably better than the traditional threshold function . By using Matlab software platform , our algorithm implementation. The simulation results show that our proposed enhancement algorithm based on human auditory characteristics of wavelet packet transform voice in both subjective and objective achieved better enhancement effect .

Related Dissertations

  1. The Research of Speech Coding Based on Compressed Sensing,TN912.3
  2. Multi- directional tube array speech enhancement system development,TN912.35
  3. Windowed Lifting Wavelet Packet in Fault Location for Research,TM755
  4. Complex environment speech enhancement method,TN912.35
  5. Research on Electromagnetic Emission Method Based on Coal and Gas Outburst Prediction,TD713
  6. Research on Image Denoising with Wavelet Transform,TP391.41
  7. Research on Key Technologies of Digital Hearing Aids Based on Auditory Masking,TN912.3
  8. The Research and Implementation of Speech Enhancement Algorithm Based on Bark Domain,TN912.35
  9. Study on the De_noising of Partial Discharge Signal Based on Wavelet Transform,TM83
  10. Time Series Analysis Using Gaussian Process,O211.61
  11. Real-time speech enhancement and its DSP system implementation,TN912.3
  12. The Study of Wavelet Packet Analysis for Denoising and the Application on GPS Data Process,TP391.41
  13. Study on the Design of Intelligent Test System of Electromagnetic Interference,TP274.4
  14. The Development of MPEG-4 Audio Coding System Based on Wavelet Packet Decomposing Algorithm,TN912.3
  15. Image Denoising Method Based on Wavelet Transform,TP391.41
  16. The Study of ECG Feature Extraction Method Based on Wavelet (Packet) Transformation,TP391.41
  17. Research on Speech Recognition in Noisy Environment,TN912.34
  18. Sample Synchronization and Blind Channel Estimation for Multi-Carrier Modulation Based on Wavelet Packet Transform,TN911.3
  19. Study on Microphone Array Speech Enhancement Algorithm,TN912.3
  20. The Research of Speaker Recognition under Noisy Environment,TP391.42
  21. Array speech enhancement algorithm and,TP301.6

CLC: > Industrial Technology > Radio electronics, telecommunications technology > Communicate > Electro-acoustic technology and speech signal processing > Speech Signal Processing > Speech Enhancement
© 2012 www.DissertationTopic.Net  Mobile