MultiCam - Low Cost Multispectral Camera (POCI-01-0247-FEDER-69271)

The present project aims to develop a new low-cost multispectral camera that can be left in the field, sending images regularly to a user. The multispectral camera will be formed by an array of cameras each with a transparent optical filter at a certain wavelength of light. Two uses of multispectral technology will be developed: 1) Automatic identification system for fish in fishing vessels; 2) Non-destructive measurement of sugars in grapes. This will be done using artificial intelligence algorithms and machine learning. Fish identification will be done by collecting image data at equally spaced wavelengths. The identification of fish species will aim to estimate the amount of fish caught by species. For measuring sugars in grapes we will make a selection of the most suitable wavelengths and we will use these wavelengths to create a spectral index that has a high correlation with the amount of sugar. Finally, we will connect our multispectral camera to the XtraN platform for managing freight transport fleets and to the Monicap box used in the fisheries sector, which is highly innovative.