Research And Implementation Of Voice Laser Modulation Signal Enhancement Technology Based On Amplitude Spectrum Estimation

Speech laser modulation signal enhancement processing is one of the key technologies of speech laser modulation system.This paper is based on the research project of the voice laser modulation system,and the enhancement processing of the voice laser modulation signal,which belongs to the software development part of the whole machine project.The source and characteristics of noise are deeply studied and analyzed,and the noise suppression method is designed to deal with specific noise.For the broadband background noise,the endpoint detection algorithm based on spectral entropy method is improved to improve the detection accuracy of the algorithm,and the pure noise and noisy speech signals are separated.In order to obtain a speech enhancement algorithm which is effective at low SNR,the speech enhancement algorithm based on spectral subtraction is improved,and the simulation experiment of speech laser modulation signal enhancement algorithm based on spectral subtraction is completed.The simulation results show that the improved algorithm has better enhancement under low SNR.The experimental platform of the speech laser modulation system was built,and the experiment was carried out in conjunction with the hardware system.The speech laser modulation signal enhancement system was experimentally verified.The speech laser modulation signal enhancement technology is mainly divided into three parts: theoretical analysis and demonstration,algorithm design and simulation and experimental verification to give the evaluation results.In the theoretical analysis and argumentation,the characteristics of speech laser modulation signals are analyzed from three aspects: time domain,frequency domain and spectral map.The mathematical modeling of speech signal generation process is carried out.The sources of speech laser modulation signal noise are analyzed,including impulse noise,power frequency interference,building noise,hardware equipment noise,wideband background noise,and some suppression algorithms are proposed based on the characteristics of noise.In the algorithm design and simulation part,according to the analysis of speech laser modulation signal and noise signal in the theoretical analysis stage,the algorithm of speech laser modulation signal endpoint detection based on spectral entropy under low SNR is proposed.The noisy speech signal and pure noise are applied.The signal is separated.The simulation experiment of the endpoint detection algorithm with low SNR is proposed.The simulation results show that the proposed detection algorithm with low SNR has a high detection accuracy under low SNR.The characteristics of wideband background noise of speech laser modulated signals are analyzed.The speech laser modulation signal enhancement technology based on spectral subtraction is improved and verified by simulation experiments.The simulation results show that the improved spectral subtraction has a good inhibitory effect on the broadband background noise of the speech laser modulated signal.In the experimental verification part,the experimental platform of the speech laser modulation system was built,and the hardware verification system was used to verify the experiment.The speech laser modulation signal was collected by the hardware experiment platform and processed by the speech laser modulation signal enhancement algorithm in the thesis.The experimental results show that the speech laser modulation signal enhancement technique proposed in this thesis can effectively improve the quality of speech laser modulation signals.