Flexible vitality of near-infrared spectroscopy – Talking about Norris derivative filter

Abstract

Near-infrared spectroscopic analysis is a modeling system that fuses three multidimensional spaces of samples, variables, and models. Spectral preprocessing is a basic skill that plays an important role in information extraction, de-noising, model maintenance, and transfer. Due to the variety of objects, conditions, and measurement methods, pre-processing patterns often need to meet personalization requirements. Norris derivative filter adopts three variable parameters: the derivative order: d, the number of smoothing points: s and the number of differential gaps: g, which is a multi-mode algorithm group. The fusion of parameters with different functions can enhance the flexible vitality of near-infrared spectroscopy and meet the personalized needs of the spectral preprocessing diversity. In this study, a partial least squares (PLS)-based platform was constructed to achieve global optimal selection of Norris derivative filter algorithm. Taking the near-infrared analysis of corn crude protein as an example, we want to share the understanding of Norris derivative filter method. In the early stage of this paper production, we were shocked to hear that Dr. Karl H. Norris died! We did this study in memory of Dr. Karl H. Norris.