CAJ | 학술논문

通过分析含噪语音信号的特点，引入能够兼顾人耳听觉特性的听觉感知小波变换，构造了新的小波阈值函数，并对小波变换分解后的阈值进行基于微粒群算法的分层优化。仿真实验表明，该方法在不同信噪比条件下均具有较好的去噪性能，语音的可懂度和听觉效果得到有效提高。
통과분석함조어음신호적특점，인입능구겸고인이은각특성적은각감지소파변환，구조료신적소파역치함수，병대소파변환분해후적역치진행기우미립군산법적분층우화。방진실험표명，해방법재불동신조비조건하균구유교호적거조성능，어음적가동도화은각효과득도유효제고。
By analyzing the characteristic of noisy speech signals, audio perception wavelet transform was introduced, which considered human auditory effect. New wavelet threshold function was constructed and hierarchical optimization was performed based on particle swarm optimization algorithm after wavelet transform. Simulation indicated that the proposed method had a good de-noising effect under circumstances of different signal-noise-ratio(SNR) , improved speech intelligibility and auditory effect.