Hyperspectral Imaging Coupled with Multivariate Analyses for Efficient Prediction of Chemical, Biological and Physical Properties of Seafood Products

Abstract

Quality evaluation of seafood is essential for consumer satisfaction. Hyperspectral imaging (HSI) has been introduced in the seafood industry for assessing seafood quality, safety, authenticity, and adulteration whilst maintaining sample integrity. However, there is limited information on multivariate analyses applied using the HSI for seafood quality. This review presents a comprehensive summary of the existing published research to describe the application of HSI coupled with multivariate analyses of seafood products. Applications of multivariate analyses for map distribution, spectral selection, and data extraction of the HSI system in the seafood industry are highlighted. Trends and challenges using HSI in the seafood industry are also discussed in this review. As a rapid and non-destructive tool, HSI technology shows great potential for evaluating the quality of seafood products by on-line or at-line detection. The ability to provide spatial and spectral information coupled with multivariate analyses makes the HSI system broadly in the seafood industry. Deep learning performed by artificial intelligence is a great solution recently for data classification of hyperspectral imaging with a shift-invariant feature of seafood products. HSI systems fitted with multivariate analyses software could be eased in the large-scale seafood industry to determine the chemical, biological, and physical quality traits of seafood products.