Agriculture Reference
In-Depth Information
with UAV systems to provide accurate, site-specific crop management. Swain et al.
(2008) investigated the estimation of rice yield and protein content using near-real-
time remote sensing images acquired by a digital Tetracam camera associated data
acquisition system mounted on a radio-controlled unmanned helicopter platform.
The determination coefficient (
r
2
) of 0.95 was obtained between the rice yield and the
NDVI and GNDVI values at booting stage, and a positive correlation with
r
2
of 0.50
was obtained for the protein content estimated in terms of total nitrogen presented
in milled rice. The results of this research show a potential of using unmanned heli-
copter-based remote sensing images for precise estimation of rice yield, which can
be useful for making important management decisions on crop cultivation. Swain
et al. (2010) developed a radio-controlled unmanned helicopter-based low-altitude
remote sensing (LARS) platform to acquire quality images of high spatial and tem-
poral resolution for estimation of yield and total biomass of a rice crop (
Oriza sativa
L.). NDVI values at panicle initiation stage were found to have high correlation with
yield and total biomass with regression coefficients (
r
2
) of 0.728 (RMSE = 0.458 ton
ha
-1
) and 0.760 (RMSE = 0.598 ton ha
-1
), respectively. The results show that LARS
images can be used to replace satellite images for estimating leaf chlorophyll content
in terms of NDVI values (
r
2
= 0.897, RMSE = 0.012). On the basis of the LARS sys-
tem, there is potential to monitor nutrients at critical growth stages in required areas
to improve final yield in rice cropping. Recently, Xiang and Tian (2011) developed
a low-cost agricultural remote sensing system based on an autonomous UAV, which
is an easily transportable helicopter platform weighing less than 14 kg. A multispec-
tral camera and autonomous system were mounted on the UAV system, making it
capable of acquiring multispectral images at the desired locations and times. Sensor
fusion techniques were used in the implementation of the UAV navigation system
that was designed using an extended Kalman filter. The interface between a human
operator and the UAV was operated based on a designed ground station to carry out
mission planning, flight command activation, and real-time flight monitoring. The
UAV could be automatically navigated to the desired waypoints on the basis of the
navigation data, and the waypoints generated by the ground station. Then the UAV
could hover around each waypoint to collect field image data. The experimental
results show that the UAV system is flexible and reliable for sensing agricultural
field with high spatial and temporal resolution of image data. Although UAVs are not
inconvenient to carry out on long flights because of the payload and engine opera-
tion time, we envision that small, lightweight UAVs will be available to satisfy the
needs of resource management agencies, rangeland consultants, and private land
managers for acquiring improved data at a reasonable cost, and for making appro-
priate management decisions in the future. Moreover, as two main remote sensing
technologies, both satellite remote sensing and aerial photography have their own
advantages and constraints. Several studies have compared satellite remote sensing
and aerial photography in terms of their ability to estimate nutrition information of
crops. Han et al. (2001) compared SPOT satellite images and digital aerial images
for their ability to detect in-season nitrogen stress in corn on two commercial fields
in 1999 and 2000. The results show that when the field had large spatial variability in
crop development, there were strong correlations between spectral variables derived
from aerial images and those from SPOT images. However, when the crop was more
