Date of Award
8-2024
Document Type
Thesis
Degree Name
Master of Science (MS)
Department
Agricultural Education
Committee Chair/Advisor
Bulent Koc
Committee Member
Matias Jose Aguerre
Committee Member
John P. Chastain
Abstract
Pastures cover a major part of the agricultural land in the United States. These prominent land covers need finer maintenance to boost biomass production and optimize consumption. A vital factor in this process is the precise knowledge of the forage availability. This study focused on the development of sensor-based systems for estimating the aboveground biomass in pastures. The study was conducted on Alfalfa and Bermudagrass, which are the widely grown legumes and grasses in the country for grazing and hay production. This study compared the performance of five different crop height measurement systems and a vegetation coverage measurement system for predicting the wet and dry biomass yield in these crops. The Structure-from-Motion (UAV-SfM), Ultrasound Sensor and Ski (US-Ski), Inertial Measurement Unit and Ski (IMU-Ski), IMU and Roller (IMU-Roller) and a Depth Camera (DC) were the five systems used in this study. For Alfalfa, the results indicated that the UAV-based Structure-from-Motion (UAV-SfM) and the Inertial Measurement Unit and Ski (IMU-Ski) based systems outperformed others. The UAV-SfM produced an R2 of 0.74 with a SeY of 2543 kg-wet/ha, and the IMU-Ski produced an R2 and SeY of 0.79 and 3166 kg-wet/ha, respectively. For the Bermudagrass, the Depth Camera (DC), the IMU-Ski, and the Ultrasound Sensor and Compression Ski (US-Ski) systems showed the promising results. Out of these three systems, the IMU-Ski was the best (R2 = 0.97; SeY = 1112 kg-wet/ha), followed by DC and US-Ski as the second and third best system for Bermudagrass respectively. The vegetation coverage (VC) itself produced satisfactory results as an independent variable for wet biomass estimation in Bermudagrass but did not perform well for Alfalfa. Additionally, the integration of VC with the crop height-based measurements did not improve the model’s predictions for Alfalfa and Bermudagrass. Thus, for both crops in this study, the method based on the IMU-Ski measurements was the recommended model for wet and dry biomass predictions.
Recommended Citation
Singh, Jasanmol, "Aboveground Biomass Estimation of Pastures Using Field Robotics" (2024). All Theses. 4347.
https://open.clemson.edu/all_theses/4347
Author ORCID Identifier
0000-0001-8893-7030
Comments
Degree Program: MS in Agriculture (Agricultural Systems Management)
Department of Agricultural Sciences, Clemson University, Clemson, SC-29630