Discrimination of Different Varieties of Large Indian Cardamom Using NIR Spectroscopy

The paper presents a novel approach to discriminate between different varieties of Large Indian Cardamom (LIC) utilizing Near-Infrared (NIR) spectroscopy. Large Indian Cardamom, a high-value spice, is renowned for its aromatic and medicinal properties, with multiple varieties exhibiting subtle differences in flavor and bio-active compounds. Traditional methods of variety discrimination are time-consuming, labor-intensive, and often subjective. This study leverages the rapid and non-destructive capabilities of NIR spectroscopy to develop a robust discrimination model. Eight different varieties were taken and their absorbance spectra were recorded within the 900–1700 nm range. Various data preprocessing techniques, including Min–Max scaler, Robust scaler, and Standard Normal Variate (SNV) were used. Principal Component Analysis (PCA) was employed to extract meaningful information from the spectra and build a classification model. The PCA shows good discrimination between the samples, with a high separability index (SI) of 213.68. Results demonstrate the feasibility of using NIR spectroscopy to distinguish between different LIC varieties with high accuracy. The discriminatory power of the model is further enhanced by combining it with advanced machine learning algorithms, ensuring reliability in practical applications. This research not only streamlines LIC variety discrimination but also has broader implications for quality control and authentication in the spice industry. NIR spectroscopy, as a rapid and non-destructive tool, holds great promise for ensuring the authenticity and quality of agricultural products, contributing to both consumer satisfaction and industry standards.