Calibrating lab and field reflectance spectra for nutrient estimation in potato plants using local support vector regression models

Abstract

This study presents a methodology based on multiple local support vector regression (SVR) for calibrating the spectra taken in the field in relative to lab-derived spectra. Laboratory based foliar spectral measurement is a common method to provide lab-derived spectra as a service where a grower sends sample leaves collected manually. The drawback of this method is being time-consuming when the samples are collected and analyzed. In contrast, in-field spectral measurements can be an alternative method capable of providing immediate readings. This method, however, is conditional on the grower being able to obtain and operate an in-field spectroscopy instrument. While both methods use the same technology, the difference in the measurement conditions may cause differences in the results. In this work, after developing the calibration method, we validated it by estimating NPK measurements in potato plants using in-field, lab, and field calibrated spectral measurements over two testing modes: dried and fresh. The results showed that the calibration using SVR models could minimize the relative error between lab and field spectra within the visible range by considering the influence of the neighboring wavebands up to 32 nm width. Also, the calibration showed a substantial reduction in error values from 120% to 20% for some wavebands in the short-wave infrared (SWIR) region of the fresh mode. Moreover, the calibration improved the alignment of NPK estimated values between lab and field calibrated spectra of both modes with an emphasis on its necessity to estimate nutrients in the fresh mode where their significant wavebands are in the SWIR.