Portable NIR Spectroscopy Combined with Machine Learning for Kiwi Ripeness Classification: An Approach to Precision Farming

This study aims to evaluate and classify the ripening stages of yellow-fleshed kiwifruit by integrating spectral and physicochemical data collected from the pre-harvest phase through 60 days of storage. A portable near-infrared (NIR) spectrometer (900–1700 nm) was used to develop predictive models f...

Full description

Saved in:
Bibliographic Details
Main Authors: Giuseppe Altieri, Sabina Laveglia, Mahdi Rashvand, Francesco Genovese, Attilio Matera, Alba Nicoletta Mininni, Maria Calabritto, Giovanni Carlo Di Renzo
Format: Article
Language:English
Published: MDPI AG 2025-06-01
Series:Applied Sciences
Subjects:
Online Access:https://www.mdpi.com/2076-3417/15/11/6233
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850160861995335680
author Giuseppe Altieri
Sabina Laveglia
Mahdi Rashvand
Francesco Genovese
Attilio Matera
Alba Nicoletta Mininni
Maria Calabritto
Giovanni Carlo Di Renzo
author_facet Giuseppe Altieri
Sabina Laveglia
Mahdi Rashvand
Francesco Genovese
Attilio Matera
Alba Nicoletta Mininni
Maria Calabritto
Giovanni Carlo Di Renzo
author_sort Giuseppe Altieri
collection DOAJ
description This study aims to evaluate and classify the ripening stages of yellow-fleshed kiwifruit by integrating spectral and physicochemical data collected from the pre-harvest phase through 60 days of storage. A portable near-infrared (NIR) spectrometer (900–1700 nm) was used to develop predictive models for soluble solids content (SSC) and firmness (FF), testing multiple preprocessing methods within a Partial Least Squares Regression (PLSR) framework. SNV preprocessing achieved the best predictions for FF (R<sup>2</sup>P = 0.74, RMSEP = 12.342 ± 0.274 N), while the Raw-PLS model showed optimal performance for SSC (R<sup>2</sup>P = 0.93, RMSEP = 1.142 ± 0.022°Brix). SSC was more robustly predicted than FF, as reflected by RPD values of 2.6 and 1.7, respectively. For ripening stage classification, an Artificial Neural Network (ANN) outperformed other models, correctly classifying 97.8% of samples (R<sup>2</sup> = 0.95, RMSE = 0.08, MAE = 0.03). These results demonstrate the potential of combining NIR spectroscopy with AI techniques for non-destructive quality assessment and accurate ripeness discrimination. The integration of regression and classification models further supports the development of intelligent decision-support systems to optimize harvest timing and postharvest handling.
format Article
id doaj-art-0028ea3dd41449f39e24fdb6ab3cdce0
institution OA Journals
issn 2076-3417
language English
publishDate 2025-06-01
publisher MDPI AG
record_format Article
series Applied Sciences
spelling doaj-art-0028ea3dd41449f39e24fdb6ab3cdce02025-08-20T02:23:01ZengMDPI AGApplied Sciences2076-34172025-06-011511623310.3390/app15116233Portable NIR Spectroscopy Combined with Machine Learning for Kiwi Ripeness Classification: An Approach to Precision FarmingGiuseppe Altieri0Sabina Laveglia1Mahdi Rashvand2Francesco Genovese3Attilio Matera4Alba Nicoletta Mininni5Maria Calabritto6Giovanni Carlo Di Renzo7Department of Agricultural, Forestry, Food and Environmental Sciences (DAFE), University of Basilicata, 85100 Potenza, ItalyDepartment of Agricultural, Forestry, Food and Environmental Sciences (DAFE), University of Basilicata, 85100 Potenza, ItalyCentre for Business and Industry Transformation, Nottingham Trent University, Nottingham NG1 4FQ, UKDepartment of Agricultural, Forestry, Food and Environmental Sciences (DAFE), University of Basilicata, 85100 Potenza, ItalyDepartment of Agricultural, Forestry, Food and Environmental Sciences (DAFE), University of Basilicata, 85100 Potenza, ItalyDepartment of Agricultural, Forestry, Food and Environmental Sciences (DAFE), University of Basilicata, 85100 Potenza, ItalyDepartment of Agricultural, Forestry, Food and Environmental Sciences (DAFE), University of Basilicata, 85100 Potenza, ItalyDepartment of Agricultural, Forestry, Food and Environmental Sciences (DAFE), University of Basilicata, 85100 Potenza, ItalyThis study aims to evaluate and classify the ripening stages of yellow-fleshed kiwifruit by integrating spectral and physicochemical data collected from the pre-harvest phase through 60 days of storage. A portable near-infrared (NIR) spectrometer (900–1700 nm) was used to develop predictive models for soluble solids content (SSC) and firmness (FF), testing multiple preprocessing methods within a Partial Least Squares Regression (PLSR) framework. SNV preprocessing achieved the best predictions for FF (R<sup>2</sup>P = 0.74, RMSEP = 12.342 ± 0.274 N), while the Raw-PLS model showed optimal performance for SSC (R<sup>2</sup>P = 0.93, RMSEP = 1.142 ± 0.022°Brix). SSC was more robustly predicted than FF, as reflected by RPD values of 2.6 and 1.7, respectively. For ripening stage classification, an Artificial Neural Network (ANN) outperformed other models, correctly classifying 97.8% of samples (R<sup>2</sup> = 0.95, RMSE = 0.08, MAE = 0.03). These results demonstrate the potential of combining NIR spectroscopy with AI techniques for non-destructive quality assessment and accurate ripeness discrimination. The integration of regression and classification models further supports the development of intelligent decision-support systems to optimize harvest timing and postharvest handling.https://www.mdpi.com/2076-3417/15/11/6233portable near-infrared spectroscopyharvest quality predictionfruit ripening classificationmachine learning model
spellingShingle Giuseppe Altieri
Sabina Laveglia
Mahdi Rashvand
Francesco Genovese
Attilio Matera
Alba Nicoletta Mininni
Maria Calabritto
Giovanni Carlo Di Renzo
Portable NIR Spectroscopy Combined with Machine Learning for Kiwi Ripeness Classification: An Approach to Precision Farming
Applied Sciences
portable near-infrared spectroscopy
harvest quality prediction
fruit ripening classification
machine learning model
title Portable NIR Spectroscopy Combined with Machine Learning for Kiwi Ripeness Classification: An Approach to Precision Farming
title_full Portable NIR Spectroscopy Combined with Machine Learning for Kiwi Ripeness Classification: An Approach to Precision Farming
title_fullStr Portable NIR Spectroscopy Combined with Machine Learning for Kiwi Ripeness Classification: An Approach to Precision Farming
title_full_unstemmed Portable NIR Spectroscopy Combined with Machine Learning for Kiwi Ripeness Classification: An Approach to Precision Farming
title_short Portable NIR Spectroscopy Combined with Machine Learning for Kiwi Ripeness Classification: An Approach to Precision Farming
title_sort portable nir spectroscopy combined with machine learning for kiwi ripeness classification an approach to precision farming
topic portable near-infrared spectroscopy
harvest quality prediction
fruit ripening classification
machine learning model
url https://www.mdpi.com/2076-3417/15/11/6233
work_keys_str_mv AT giuseppealtieri portablenirspectroscopycombinedwithmachinelearningforkiwiripenessclassificationanapproachtoprecisionfarming
AT sabinalaveglia portablenirspectroscopycombinedwithmachinelearningforkiwiripenessclassificationanapproachtoprecisionfarming
AT mahdirashvand portablenirspectroscopycombinedwithmachinelearningforkiwiripenessclassificationanapproachtoprecisionfarming
AT francescogenovese portablenirspectroscopycombinedwithmachinelearningforkiwiripenessclassificationanapproachtoprecisionfarming
AT attiliomatera portablenirspectroscopycombinedwithmachinelearningforkiwiripenessclassificationanapproachtoprecisionfarming
AT albanicolettamininni portablenirspectroscopycombinedwithmachinelearningforkiwiripenessclassificationanapproachtoprecisionfarming
AT mariacalabritto portablenirspectroscopycombinedwithmachinelearningforkiwiripenessclassificationanapproachtoprecisionfarming
AT giovannicarlodirenzo portablenirspectroscopycombinedwithmachinelearningforkiwiripenessclassificationanapproachtoprecisionfarming