【摘要】 本文结合经典Donoho小波消噪模型,提出了基于高阶统计理论的小波消噪算法。算法在小波分解各尺度引入了高阶统计理论中的四阶累积量方法,从信号和噪声的统计特性出发进行信噪分离。仿真结果验证了本文算法的性能。算法能满足无线电监测系统中对信号进行后续处理的需求。更多还原

【Abstract】 Based on the Donoho denoising model,the paper presents a wavelet-based denoising algorithm with the fourth-order cumulant in Higher Order Statistics(HOS).The proposed algorithm introduces the fourth-order analysis method in HOS across all wavelet scales,and detects the signal coefficients from the noise by their different statistics characters.Simulation results show that the proposed algorithm has a remarkable superior ability to denoise.The algorithm can satisfy the demand of signal processing in radio inspection system.更多还原