Near-infrared soil spectrum denoising method based on self-adapting filtering

Abstract

A near-infrared soil spectrum denoising method based on self-adapting filtering relates to the field of spectrum pre-processing based on self-adapting filtering, and the invention aims at solving the problems of denoising distortion, complex calculation and slow calculating speed in the prior art and can effectively inhibit the noise in the directly collected near-infrared soil spectrum. The near-infrared soil spectrum denoising method comprises the steps of: utilizing a near-infrared spectrograph to collect the soil spectrum data at any moment t1 as an input signal, collecting the soil spectrum data at a t2 moment as a reference signal in the same soil sample and on the conditions of invariable test conditions and soil ingredient contents, filtering the reference signal through a self-adapting filter to obtain an output signal, and continuously adjusting the filtering parameter according to the difference value e(lambda) of the output signal and the input signal through a normalized least mean square error self-adapting filtering algorithm to enable the e(lambda) to meet the least mean square error standard.. The near-infrared soil spectrum denoising method based on self-adapting filtering provided by the invention has no distortion, simple and convenient calculation and fast calculation speed during the denoising process.

Description

technical field [0001] The invention relates to the field of spectral preprocessing based on adaptive filtering, and is particularly suitable for spectral denoising in near-infrared soil component analysis. Background technique [0002] The adaptive filtering method was proposed by B. Widrow et al. on the basis of Kalman filtering in the 1960s. It has the advantages of small amount of calculation, fast speed, and recursive real-time processing of Kalman filtering. Based on this Improvements are made on the above, so that it does not require the statistical characteristics of the known signal, and the optimal filtering can also be achieved by automatically adjusting its own parameters. In the analysis process of near-infrared soil spectrum, the spectrum directly obtained by the spectrometer is often interfered by factors such as temperature, humidity, instrument instability, resistance heating effect, etc., so that the obtained soil spectrum contains noise, which eventually l...

Images