Development of a cost-effective hyperspectral camera for food quality monitoring

Hyperspectral imaging (HSI) is widely used for the quality assessment of various agrifood products thanks to the capability to retrieve useful information (both spatial and spectral) which can be qualitatively and/or quantitatively modelled for the prediction of several physical and chemical attributes. Even though HSI technique can collect a large amount of data, the application of only one device (in some cases) is not enough to cover all the critical points that the company has to handle. Both the production process in the firm and the field monitoring could require distributed systems to collect data and provide information. Under these circumstances, considering the application of several hyperspectral devices, the costs become prohibitive for most of the companies and, therefore, the research is moving toward the development of hyperspectral solutions taking into account a considerable cost reduction for distributed stand-alone applications. Therefore, a prototype of a HSI camera has been designed and built starting from a previous work proposed by Salazar and Mendez (2020). The final system showed the capability to pick the spectral information from 400-1000 nm within a spatial surface of 41x41 pixels. The whole system has been fine-tuned using low-cost optical and electronic components (i.e., lenses, transmissive holographic diffraction gratings, Raspberry Pi 3b+ and a Raspberry Pi NoIr camera) and parts 3D printed using a photopolymer resin. A first trial has been performed in lab scale conditions in order to verify the sensibility and sensitivity of the HSI device to acquire hyperspectral images which can be handled and analysed in MATLAB environment.