ISHS Acta Horticulturae 674: III International Symposium on Applications of Modelling as an Innovative Technology in the Agri-Food Chain; MODEL-IT
MULTIVARIATE ANALYSIS OF AN ON-LINE NIR SPECTROMETER UNDER INDUSTRIAL USE |
| Authors: | P. Barreiro, M. Ruiz-Altisent, C. Bielza, A. Moya-González |
| Keywords: | robustness analysis, NIR application, onion, breeding, classification |
| Abstract:
This study validates an unsupervised procedure for the identification of daily event (general changes) and abnormal observations for an on-line NIR spectrometer under industrial use. Process control statistics (Hotelling T2, Q) are used for a multivariate supervision of the onion bulb classification under breeding strategy. Since interactance is used for sample presentation, real time detection of abnormal spectra avoids misclassifications due to poor contact between bifurcated fibber and bulbs as well as to faults in the behaviour of the equipment which eventually occur due to the aggressiveness of the environment. | |
Download Adobe Acrobat Reader (free software to read PDF files) | |
