基于小波变换的昆虫刺吸电位(EPG)信号去噪研究

Study of Insect Electrical Penetration Graph( EPG) Signal Denoising Based on Wavelet Transform

昆虫刺吸电位(EPG)信号为研究刺吸式昆虫取食和传毒机理提供了有力的依据,然而EPG信号在采集过程中易受各种噪声的干扰,可利用小波变换将其去除。本文采用小波阈值去噪法,通过实验对比选择不同的小波基函数及阈值,以均方根误差和信噪比为评价指标,确定了coif4小波6层分解,在Stein无偏风险估计准则下,改进阈值量化为最佳去噪方案;实验结果表明,该小波阈值去噪方法可以有效地去除不同干扰。本文研究的基于小波变换的EPG信号去噪为后续分析识别提供了保证。

Electrical penetration graph（ EPG） signal has provided a strong basis for the study of the piercing-sucking insects feeding and virus transmission mechanism. However,the EPG signal is susceptible to various noises during acquisition,which can be removed by wavelet transform. In this paper,the wavelet threshold denoising method was adopted. The different wavelet basis functions and thresholds of the denoising method were selected by experiments.Taking the root mean square error and signal-to-noise ratio as the evaluation index,the wavelet coif4,6 layer decomposition and the improved threshold quantization was determined as the best denoising scheme under the Stein unbiased risk estimator. The experimental results showed that the wavelet threshold denoising method can effectively remove different interference. The EPG signal denoising based on wavelet transform in this paper provides a guarantee for subsequent analysis and recognition.