The impact of high-quality data on the assessment results of visible/near-infrared hyperspectral imaging and development direction in the food fields: a review

Over the past decades, food quality and safety have seriously been disturbing public health. Traditional monitoring methods of food-quality evaluation have been time consuming and laborious. In recent years, visible/near-infrared hyperspectral imaging (Vis/NIR-HSI)—as a nondestructive, high-efficiency, and rapid method—has become one of the strongest tools for food evaluation. Using this tool involves an analysis process that includes obtaining high-quality data and training models. Although recent works have mainly focused on the development and optimization of models—ignoring the importance of high-quality data acquisition—this paper reviews the precautionary measures taken in the form of the operational steps that are followed in the process of acquiring food-related information. This includes the spectral mode selection, sample, waveband range, ROI extraction, spectra parameters, data processing, and focuses on finding solutions to problems such as large curvature and specular reflection. The application and selection of the quality parameters of different food items were introduced in their assessment. Finally, based on the above summary, an updated overview of the application of Vis/NIR-HSI have been presented and some viewpoints and directions for future research have also been provided.