Grading and color evolution of apples using RGB and hyperspectral imaging vision cameras Original Research Article

The potential of RGB digital imaging and hyperspectral imaging (900–1700 nm) was evaluated for discriminating maturity level in apples under different storage conditions along the shelf-life. Segmentation, preprocessing and partial least squares-discriminant analysis (PLS-DA) were used for hyperspectral data analysis, while illumination correction, dimensionality reduction and linear discriminant analysis (LDA) were used for RGB data analysis. The results showed that hyperspectral discrimination classified different storage regimes better than RGB, with an overall success rate of 95.83%. In addition, color evolution of apples during shelf-life under different storage regimes was modeled using RGB zero and first order regression models, fitting better to a first order kinetic model.