The dredging is a process that intrinsically damages the aquatic environment. Suctioning part of the aquatic bottom surface suppose not only change the ecosystem but it endanger the life of the animal and plant species. Nowadays, there is doing a lot of efforts to improve the ecological aspect of the dredging process. In this work, we propose the introduction of machine vision techniques to obtain this improvement, using hyperspectral imaging. The performed tests show that is possible to reduce the environmental impact applying these techniques in two points of the dredging process.