Measurement of Cotton Canopy Temperature Using Radiometric Thermal Sensor Mounted on the Unmanned Aerial Vehicle (UAV)

Juan A Landivar,Murilo M Maeda,Jinha Jung,Henrique D R Carvalho,Anjin Chang,Junho Yeom

doi:10.1155/2020/8899325

Abstract

Canopy temperature is an important variable directly linked to a plant’s water status. Recent advances in Unmanned Aerial Vehicle (UAV) and sensor technology provides a great opportunity to obtain high-quality imagery for crop monitoring and high-throughput phenotyping (HTP) applications. In this study, a UAV-based thermal system was developed to directly measure canopy temperature, skipping the traditional radiometric calibration process which is time-consuming and complicates data processing. Raw thermal imagery collected over a cotton field was converted to surface temperature using the Software Development Kit (SDK) provided by the sensor company. Canopy temperature map was generated using Structure from Motion (SfM), and Thermal Stress Index (TSI) was calculated for the test site. UAV temperature measurements were compared to ground measurements acquired by net radiometers and thermocouples. Temperature differences between UAV and ground measurements were less than 5%, and UAV measurements proved to be more stable. The proposed UAV system was successful in showing temperature differences between the cotton genotype. In conclusion, the system described in this study could possibly be used to monitor crop water status in a field setting, which should prove helpful for precision agriculture and crop research.

Highlights

Canopy temperature is an important indicator of water availability, water stress, and irrigation status in agriculture [1]
A canopy temperature map was created from thermal images collected with the Unmanned Aerial Vehicle (UAV)-based system using the Structure from Motion (SfM) algorithm and geotagged temperature images (see Figure 4(a))
An orthomosaic image with finer spatial resolution could be generated since rotary-wing UAVs can fly at lower altitude with fairly stable orientation, when compared to a fixed-wing system

Summary

Introduction

Canopy temperature is an important indicator of water availability, water stress, and irrigation status in agriculture [1]. Sensed data, which are acquired by sensors on space-borne, air-borne, or groundbased platforms, have been widely used in agriculture to estimate crop parameters such as vegetation indices and Leaf Area Index (LAI) [3, 4]. Most studies using UAV for agriculture have focused on red-green-blue (RGB) and/or multispectral sensors to calculate vegetation indices and monitor crop development for yield forecasting. Advanced UAV systems can provide fine spatial and high temporal resolution data at relatively low cost so that crop traits such as height, canopy morphology, and greenness can be estimated [2, 6,7,8]. Patrick and Li [8] generated 3D models of blueberry bushes from UAV data to extract morphological traits for genotype selection and found a strong relationship between traditional growth indices and image-derived bush volume. Zhang et al [11] developed nonlinear regression models to predict sorghum biomass from multitemporal UAV-based hyperspectral and RGB data, while Ali et al [12] adopted hyperspectral, LiDAR, and RGB data for sorghum biomass prediction

Methods

Results

Conclusion