Mango Quality Assessment Using Near-Infrared Spectroscopy and Hyperspectral Imaging: A Systematic Review

Mango is considered a high-value tropical fruit, and its commercial and consumer acceptance depends on internal and external quality attributes such as Total Soluble Solids (TSS), Dry Matter Content (DMC), firmness, ripeness, and surface defects. In recent years, non-destructive sensing technologies such as Near-Infrared Spectroscopy (NIRS) and Hyperspectral Imaging (HSI) have gained prominence for their ability to quickly and accurately evaluate mango quality. In this study, 101 articles published within the last ten years, were systematically retrieved, and 85 research papers were selected for detailed analysis. The review focuses on statistical analysis, conventional machine learning, deep learning, and transformer-based methods applied to mango quality assessment. The objective is to systematically review and analyse data-driven models for non-destructive mango grading using NIRS and HSI technologies, with particular emphasis on data collection methods, preprocessing techniques, dimensionality reduction, and predictive modelling approaches. This review aims to identify the most effective and widely adopted machine learning and deep learning methods, especially transformer models, for accurate and real-time mango quality assessment. Furthermore, it highlights key quality traits evaluated, current research gaps, and future opportunities to advance intelligent, real-time, and automated mango grading systems for practical use in the fruit industry.