Classification of Cucumber Leaves Based on Nitrogen Content Using the Hyperspectral Imaging Technique and Majority...

Area of Science:

• Agricultural Science
• Food Safety
• Spectroscopy

Background:

• Excess nitrogen in cucumber cultivation leads to nitrate accumulation, posing a risk of food poisoning.
• Non-destructive quality evaluation of fruits and vegetables is crucial for food safety.
• Hyperspectral imaging offers a promising technology for assessing produce quality.

Purpose of the Study:

• To identify excess nitrogen levels in cucumber plants using hyperspectral imaging.
• To develop a reliable classification system for detecting nitrogen-rich cucumber leaves.

Main Methods:

• Utilized hyperspectral imaging to analyze cucumber plants.
• Employed a majority voting method combining five classifiers: ANN-ICA, ANN-HS, LDA, RBF, and KNN.
• Identified optimal wavelengths (723, 781, 901 nm) using ANN-BBO for classification.

Main Results:

• The majority voting system achieved a correct classification rate of 95.55%.
• Individual classifiers showed high accuracy, with ANN-ICA (96.14%), ANN-HS (96.11%), and KNN (95.24%) performing best.
• No significant difference in precision was found between using optimal wavelengths and the full spectrum.

Conclusions:

• Hyperspectral imaging combined with a majority voting classifier system is effective for detecting excess nitrogen in cucumber plants.
• This non-destructive method can contribute to preventing food safety issues related to nitrate accumulation.
• The developed system demonstrates success in distinguishing nitrogen-rich leaves from those with standard nitrogen content.