Water Stress Detection in Indoor Farming using Thermal Infrared Imaging

Early water stress detection is crucial for increasing crop yield, optimizing agricultural practices, and improving water management. This research proposes a novel method for detecting early water stress in lettuces using Thermal Infrared (TIR) and RGB images in a controlled environment. The method initially removes the background of the lettuce canopy to eliminate the noise, then clusters the canopy's thermal data to determine temperature variation and applies a yellow pseudo-color to annotate the stressed areas in the RGB images. The logic behind the proposed method is that during water stress, the transpiration of the leaves' stomata is reduced, and the temperature of the leaves increases. The results show that the proposed method finds water stress sooner than visual symptoms expression, providing helpful information for improved crop management and better water use in agriculture.