Dissertation > Excellent graduate degree dissertation topics show

Reconstruction of Sound Pressure Field Function Based on Compressive Sampling

Author: XiaHaiYing
Tutor: ZhouLingLing
School: Shanghai Jiaotong University
Course: Circuits and Systems
Keywords: Compressive sampling Sampling nuclear The narrowband voice signal Full sound function Noise model
CLC: TN912.3
Type: Master's thesis
Year: 2011
Downloads: 37
Quote: 0
Read: Download Dissertation

Abstract


Classic Shannon sampling theory can only be accurate reconstruction of band-limited signal. In the field of signal processing applications, some special non-band-limited signal reconstruction, has attracted widespread attention. Compressive sampling theory suggests that, for uniquely determined by the parameters of the signal by the low sampling rate sampling and reconstruction to get accurate results. In this paper, the theory in the reconstruction of the acoustic pressure field in depth research and analysis, some non-band-limited signal or broadband sound field can be rebuilt in a very low sampling rate. The results of this study can be obtained compared with the traditional algorithm, more accurate sound source location and reconstruction, and can greatly reduce the sound pressure field reconstruction problem in the hardware circuit sampling rate. This conclusion can be applied to the sound source localization, the synthetic virtual sound field and get head-related functions and other occasions. Firstly, given the mathematical model of the narrow-band sound signal is weak in noisy environments. By analyzing the spectral characteristics of the signal, respectively, the complete signal reconstruction method with dual-channel reconstruction method for the reconstruction with complex amplitude of the pulse train. On this basis, the compressed sampling method for narrow-band signals in the frequency domain, and gives the algorithm of the sound source localization. Then gives the whole sound function in the free field microphone linear array and circular array sampling model. In the the faint noise environment, respectively, with a sinc function, Gauss function, as well as E-spline as the sampling nuclear-dimensional function of the whole sound compressed reconstruction of the sampling time. By analyzing the spectral characteristics of the sound function and time domain expression-dimensional sound function Shannon theorem rebuild space, as well as reconstruction of space - when the sound function parameters. In a noisy environment, signal reconstruction in white noise, narrow-band noise and band-limited noise environment. This article also describes a closed space, the general expression of the whole sound function in the time domain, and characterized in accordance with the expression in the time dimension, that when having a similar delay position of the reconstruction pulse is large or when the number of pulses, classic compression sampling algorithm performance will decline. By analyzing the classic characteristics of the reconstructed signal compression sampling method gives a more precise reconstruction of the pulse method with similar delay, and improve the algorithm reduces the search cost. Analysis time dimension in the closed space a wholly acoustic characteristics of the function in the time domain, shows a method for the reconstruction of the function. The correctness of the algorithm simulation and experimental validation of the feasibility of the algorithm. In this paper, the sampling method for narrow-band signals in the frequency domain compression applied to the sound source localization, compared with the traditional time-delay estimation method requires a lower sampling rate, at least in theory be a more accurate time delay estimation. In this paper, based on the compression function reconstruction method of sampling sound, can be achieved by a very low sampling rate, and get a more accurate reconstruction. The same time, with similar delay pulse reconstruction, the proposed algorithm has a smaller search and computational cost.

Related Dissertations

  1. Noise Analysis and Image Processing of CMOS Image Sensor,TP212
  2. EEG based on compressed sensing compressive sampling,TN911.7
  3. Research on the Space-time Correlation Characteristics of Vector Noise Field in Boundary Conditions,TB52
  4. Research and Design of High Performance Programmable Charge-pump Phase Locked Loop,TH38
  5. Classification Between PS and Stego Images,TP391.41
  6. Study of the Image Denoising Based on HVS and TV Models,TP391.41
  7. Research on Multiple Description Video Coding Based on Compressive Sampling over Wireless Network,TN919.81
  8. Research on Several Key Technologies for Distributed Video Coding,TN919.81
  9. Error Detection and Time Backup for Atomic Clock on Satallite,TM935.115
  10. Research on Rail Dynamic Vibration Absorber and Its Applications of Urban Rail Transit,U213.4
  11. RF CMOS Noise Parameters Extraction Technology and Noise Figure Automation Measurement,TN386.1
  12. The Detail Restoration and Enhancement Method of Fluid Animation,TP391.41
  13. Study of Decoding Algorithms Based on Distributed Video Coding,TN919.81
  14. The Method of Solving Mathematical Model of the Image Restoration about X-ray Diffraction Association Imaging,TP391.41
  15. Research of UHF Differential Frequency-hopping Detection Technology Based on Compressed Sampling,TN914.41
  16. The Research of CCD Noise Model for Circuit Simulation and Testing Based on Camera,TN386.5
  17. Research on Algorithms and Applications Related to H.264/AVC Residual Coding,TN919.81
  18. Study on a 120MHz Low Phase Nosie Oven Controlled Crystal Oscillator,TN752
  19. Characteristic Analysis of Noise Acoustic Field of Fiber-Optic Vector Hydrophone,TB565.1
  20. The Research and Design of Low Noise Pre-amplifier of TEM Receiver,TN722

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