PhenAI-Bot: Precision 3D Crop Phenotyping of Pepper (Capsicum annuum L.) Varieties in Greenhouse — 65p — Swarnabha Roy, Amee Parmar, Rishik Aggarwal, Stavros Kalafatis

Manual crop phenotyping is a time-consuming and labor-intensive process. The lack of low-cost, open-access 3D phenotyping tools has limited the comprehensive exploration of dynamic 3D growth in crops across all growth stages. Crop height and leaf growth rate at different stages critically impact their overall health and hence yield. This research introduces PhenAI-Bot, an autonomous robot designed to assess five phenotypic traits (plant height, canopy major and minor diameters, canopy area, and leaf count per plant) in four varieties of pepper (Black Hungarian, Hungarian Hot Wax, Poblano, and Thai Hot; Capsicum annuum L.) within greenhouse settings. The PhenAI-Bot consists of an Intel RealSense D435i RGB-D camera, mounted on an aluminum extrusion, to capture top-down images (RGB, Depth, and 3D point cloud) of plant canopies. Six plants for each variety, totaling 24 plants were used in this study. Each plant’s pot was marked with a QR code having variety and plant number information. A secondary RGB camera mounted on the side of the PhenAI-Bot was used to scan the QR codes to autonomously retrieve variety and plant number data corresponding to the measured phenotyping traits and image data. The measured values were compared with ground truth data and Pearson’s correlation coefficient (R²) values ranged from as low as 0.015 for Hungarian Hot Wax to as high as 0.91 for Poblano. Similarly, the mean squared error (MSE) ranged from as low as 0.004 for Hungarian Hot Wax to as high as 0.097 for Thai Hot. The results demonstrate that PhenAI-Bot along with the developed method can measure phenotypic traits at a moving speed of up to 0.2 m/s, delivering acceptable accuracy under standard daytime conditions. Future work will incorporate AI algorithms to improve the measurement accuracy of phenotyping traits and expand the system to other crops, like wheat, soybean, corn, and sunflower.

South Dakota State University
Texas A&M University
Dr. Pappu Kumar Yadav