A novel non-integer order Savitzky-Golay derivative function of visible and near-infrared spectra for improving prediction accuracy of phosphorus in pig manure
- Authors
- Publication Date
- Jan 01, 2023
- Source
- Ghent University Institutional Archive
- Keywords
- Language
- English
- License
- Unknown
- External links
Abstract
Prediction of phosphorus (P) content in manure using visible and near infrared (vis-NIR) spectroscopy is essential for better manure management. However, accurate prediction requires better spectra pre-processing, out of which spectra derivative is frequently used. This study presents the development of a non-integer order SavitzkyGolay derivative function (NISGDF) of vis-NIR spectra for the analysis of P composition in manure. A formula for a NISGDF is developed and presented in the form of a matrix. Next, both quantitative and qualitative analyses were performed using developed software packages, all of which can be easily called by a command line. Simulated absorption bands using the Gaussian model were operated by NISGDF to assess its abilities in reduction of noise and elimination of baseline and offset. A total of 110 pig manure samples were collected and scanned by a fibre type vis-NIR spectrophotometer in the range of 305-1700 nm. The collected manure spectra were subjected to the NISGDF, before they were used to establish the partial least squares regression (PLSR) model for P. Results indicate that the NISGDF offers significant flexibility in interpolating between the integer order SG derivatives, and reduces baseline offsets and tilts compared to conventional integer order SavitzkyGolay derivative function (ISGDF). The PLSR prediction accuracy with NISGDF was improved by 5-8%, compared with the corresponding PLSR model using the traditional ISGDF. This work shows that the NISGDF offers the advantage of wider applicability and better performance for the prediction of manure P, hence, it is recommended as a more general and better algorithm than the traditional ISGDF for the prediction of P.
