A fast analysis approach for crop health monitoring in hydroponic farms using hyperspectral imaging

Abstract

Hydroponic farming is considered as a more sustainable solution in comparison to conventional farming. Most of the hydroponic farms rely on manual visual inspection for crop monitoring, which can be subjective, time-consuming, and tedious, especially in the case of large area farms. Hyperspectral imaging (HSI) is a promising technique for automated sensing and monitoring. Though several automated systems based on HSI have been developed recently for crop monitoring, these tend to be computationally complex and demand significant processing power and time, especially when handling extensive data from large farms. In this context, we explore an approach using spectral ratios for crop growth monitoring and the detection of early-stage nutrient stress. The early detection of the nutrient stress can enable effective crop, resource, and time management in large hydroponic farms. A sensitive nutrient deficiency index, named normalized nutrient deficiency index (NNDI), has been formulated for the early-stage detection of nutrient deficiencies. Evaluating these indices is computationally simple and quick. A methodology for crop growth monitoring and nutrient deficiency stress using these indices is demonstrated on Lactuca sativa L. crops. It is envisaged that the proposed quick, nondestructive imaging technique can enable future automation possibilities and serve as an invaluable tool in indoor hydroponic farms.