=Paper=
{{Paper
|id=Vol-1498/HAICTA_2015_paper83
|storemode=property
|title=The Use of Unmanned Aerial Systems (UAS) in Agriculture
|pdfUrl=https://ceur-ws.org/Vol-1498/HAICTA_2015_paper83.pdf
|volume=Vol-1498
|dblpUrl=https://dblp.org/rec/conf/haicta/SimelliT15
}}
==The Use of Unmanned Aerial Systems (UAS) in Agriculture==
https://ceur-ws.org/Vol-1498/HAICTA_2015_paper83.pdf
The Use of Unmanned Aerial Systems (UAS) in
Agriculture
Ioanna Simelli1, Apostolos Tsagaris2
1
MIS, University of Macedonia, Thessaloniki, Greece, e-mail: ioanna.simeli@gmail.com
2
Department of Automation, Alexander Technological Educational Institute of Thessaloniki,
Greece, e-mail: tsagaris@autom.teithe.gr
Abstract. Unmanned aerial vehicles (UAVs) represent technological
developments used for precision agriculture. They provide high-resolution
images taken from crops and when specific indices are applied, useful outputs
for farm management decision-making are produced. The current paper
provides a literature review on the use of UAVs in agriculture and specific
applications are presented.
Keywords: Precision agriculture, UAS, unmanned aerial vehicles.
1 Introduction
Unmanned Aerial Systems (UAS) are aerial vehicles, which come in wide varieties,
shapes and sizes and can be remotely controlled or can fly autonomously through
software-controlled flight plans in their embedded systems working on the basis of
GPS.
A UAS is made up of light composite materials to reduce weight and increase
position-changing capability. Due to the usage of composite material strength they
may fly at extremely high altitudes. They may have embedded various navigation
systems or recording devices such as RGB cameras, infrared cameras, etc.
Some of the advances of the use of UAS are that they are lightweight and easy to
transport, they capture high resolution and low cost images, they can fly at variety of
altitudes depending on data collection needs, they can travel areas which are not
accessible via car, boat, etc., they are extensively used in rescue operations, helping
in delivering medicines and food, providing the live status of affected area,
communicating in crisis, etc, quick availability of raw data.
2 Unmanned Aerial Systems
In recent decades, there have been significant efforts around increasing flight
duration, the payload, and tolerance to various weather conditions, resulting in
different UAV configurations with different sizes, duration of autonomy and
730
competencies. A key criterion currently used to distinguish among aircrafts is the size
and flight duration:
• Large autonomy, high altitude (HALE) UAV, such as for example,
Northrop- Grumman Ryan 's Global Hawks (65.000 feet altitude, flight
time 35 hours, payload 1,900 lbs).
• Medium Altitude Long Endurance (MALE) UAV, such as the General
Atomics Predator (27.000 feet high, 30/40 hour flight, and beneficial load
450 lbs).
• Regular use UAVs as Hunter, Shadow 200, and Pioneer (15.000 feet high,
Flight time is 5-6 hours, and 25 kg payload) and
• Small and portable UAVs from a man as the Pointer / Ranen
(AeroVironment), Janelin (BAL) or Black Pack Mini (Mission
Technologies).
3 Software and Hardware Technologies
The technologies developed for the UAV are specific in the sense that in order to
compensate for the absence of the pilot and thus enable the flight of unmanned
vehicles and their autonomous behavior, they are mainly based on the technologies of
sensors and microcontrollers, of communication systems of Ground Control Stations
and of UAS intelligence.
A very important issue in UAS is the automation system, which is used to control
the machine. This system is separated into two parts. From the one side are the
Control systems for the machine and in most cases they can be autopilot systems,
which are used to control flights with several characteristics. Such systems contain
GPS waypoint navigation with altitude and airspeed, fully integrated multi-axis
gyroscopes and accelometers, GPS systems, pressure indicators and meters, pressure
airspeed sensors, etc. All these sensors are mounted on hardware circuit boards. They
have completely independent operation including autonomous take off and landing
and Fail-safe commands programmed into the fly control system to address loss of
altitude, loss of GPS connection, or loss of modem communication. The autopilot
recognizes problems and initiates the land command, so that the UAS immediately
flies back to the start point. The UAS can also be controlled manually with a lot of
controlling systems through wireless communication.
On the other side are the control systems for the communication with the
computer. The UAS have Ground Control Software, which provides the interface
between the UAS and computer. This software enables programming of flight
patterns and their pre-flight simulation, selection of flying files and transfer to other
systems, tracking the flight path and monitoring the conditions during the flight with
a computer-UAS online flight communication system. Finally a log file is available,
with all the flight information, also available once the flight is completed.
Christophersen et al (2004) provide a small guidance, navigation, and control
system based on FPGA (Field Programmable Gate Array) and DSP (Digital Signal
Processor) technology to satisfy the requirements for more advanced vehicle
behavior in a small package. Including those two processors into the system enables
731
custom vehicle interfacing and fast sequential processing of high-level control
algorithms (Christophersen, H., et al.,, 2004). Primicerio et al. (2012) provide a UAV
platform ‘‘VIPtero’’ which is an open-source project that is available with pre-assembled
flight and brushless control boards, the main board responsible for the actual flight of
the mikrokopter is built around an ATmega1284P microcontroller (Atmel
Corporation, San Jose, CA, USA) and communicates to the six brushless controllers
via a bi-directional two-wire serial bus. An additional navigation control board
(NaviCTRL) equipped with an ARM9 microcontroller (Atmel Corporation, San Jose,
CA, USA) and MicroSD card socket for waypoint navigation data storage is also
present (Primicerio, J., et al., 2012). Pankaj Maurya (2015) uses a single board
computer system development on Soekris net4521. It is based on a 133 Mhz 486
class processor. It uses a Compact Flash module for program and data storage. The
Soekris net4521 a special electronic circuit controlled via linux platform. The data
acquisition and filtering is done on an FPGA. The sensor data is digitized using an
Analog to Digital converter card and then fed to the FPGA. The FPGA runs Kalman
filtering on the received data. It provides the filtered values of the physical quantities
– Velocity, Angle, angular velocity and the navigation quantities in the three axis
(Maurya, 2015)
In many cases the construction of the UAS is integrated with Wireless Sensor
Networks (WSN). The information retrieved by the WSN allows the UAS to
optimize their use. For example to confine its spraying of chemicals to strictly
designated areas. Since there are sudden and frequent changes in environmental
conditions the control loop must be able to react as quickly as possible. The
integration with WSN can help in that direction (Costa, F.G., et al., 2012).
5 Use of UAS in Agriculture
Agriculture couldn't be left out of the technological advances taking place
worldwide in any scientific field. Furthermore, the need to secure food and water
supplies for a global population that grows rapidly is a challenge to be addressed
using information technology. Unmanned aerial vehicles (UAVs) represent these
technological developments used for precision agriculture. They were initially used
for chemical spraying while they were the solution to visibility problems due to
cloudy weather or inaccessibility to a field of tall crops, like maize (Sugiura,
Noguchi, & Ishii, 2005). They also have the strong advantage compared to satellite
and airborne sensors of high image resolution (Jannoura, R., et al., 2015)
The production's increase, the improvement of the efficiency, the enhancement of
profitability, the reduction of environmental impacts and the availability of
quantifiable data from large farms are some of the benefits that precision agriculture
using UAVs provides (Herwitz, et al., 2004, Xiang & Tian, 2011).
Questions, however, have risen regarding the usefulness of UAVs in agriculture.
Questions regarding their effectiveness as far as the pictures taken are concerned, the
inability of UAVs to fly in diverse weather conditions like rain that affects the
quality of images or high wind, or finally the price of data elaboration.
732
It is currently financially viable for a farm manager to purchase a drone, when in
2005 it cost as much as an 120kW tractor (Sugiura, Noguchi, & Ishii, 2005).
However, the purchase price is the least of the problems since the cost of image
processing software to produce maps is far bigger. As the cost of purchasing and
utilizing UAVs within the agriculture industry falls, interest in the sector is rapidly
increasing.
5.1 Indices used in UAVs in agriculture research
Vegetation indices in remote sensing of crop weed plants is very common. Some
indices use only the red, green and blue spectral bands (Meyer & Neto, 2008). The
most common indices used are:
Green - Red Ratio Vegetation Index (GRVI) or Normalised Green - Red
Difference Index (NGRDI): Reflectance in the green and red parts of the spectrum
Leaf Area Index (LAI): It characterizes plant canopies. One-sided green leaf area
per unit gound surface area (LAI = leaf area / ground area, m2 / m2)
Normalised Difference Vegetation Index (NDVI): Ratio of the reflectance in the
near-infrared and red portions of the electromagnetic spectrum NDI = G − R / G + R
Visible Vegetation Index (VVI) provides a measure of the amount of vegetation or
greenness of an image using only information from the visible spectrum. The VVI is
given by
where R, G, and B are the red, green, and blue components of the image,
respectively, RGBo is vector of the reference green color, and w is a weight exponent
to adjust the sensitivity of the scale.
Excess Green Index (ExG): Provides a near - binary intensity image outlining a
plant region of interest (ExG = 2g − r− b)
5.2 Case studies
Coffee plantations were traditionally small (<50ha) and hand picking was the
standard harvesting procedure. However, the plantations grew enormously,
exceeding 200ha, and mechanical harvesting is currently used. Herwitz et al., used in
2004 the NASA's Pathfinder - Plus UAV as an image collection platform, with
multispectral and hyperspectral digital imagers over land areas and coastal zone
waters, for the Kauai Coffee Company, the largest coffee plantation in the US
(approx. 1400ha).
Authors used the data gathered to spot differences in overall ground cover within
fields. The acquired data revealed a positive relationship between brightness of the
coffee tree canopy and the harvested yield of ripe coffee cherries.
In 2005, Sugiura et al, used an unmanned helicopter, flying over a sugar beet field
733
and a corn field, where they adopted a real-time kinematic global positioning system,
an inertial sensor (INS) and a geomagnetic direction sensor (GDS) to acquire the leaf
area index (LAI), an important value when estimating the crop growth. To evaluate
the crop status using LAI, the accurate segmentation of crop and soil area is needed.
An autonomous UAV-based agricultural remote sensing system was used to
monitor turf grass glyphosate application (Xiang & Tian, 2011)
Detection of the vegetation in herbaceous crops is the initial important stage when
precision agriculture is applied. Thus high resolution images (mm or very few cm) is
highly required. UAVs is the perfect tool for such a mission (Torres-Sánchez, J.,
López-Granados, F., & Peña, J.M., 2015).
Torres-Sanchez et al. in 2015 used two cameras (a conventional visible camera
and a multispectral camera) on a UAV. Then they used the software eCognition
Developer 8.9 which offers various options related to Object Based Image Analysis
(OBIA) based on the Otsu's method to detect vegetation in fields of three different
herbaceous crops (maize, sunflower and wheat). They initially used the
multiresolution segmentation algorithm (MRSA). The two vegetation indices: Excess
Green (ExG) index and Normalized Difference Vegetation Index (NDVI) were used.
The plants were in their early growth stages that corresponds to the principal stage 1
(leaf development) of the ‘‘Biologische Bundesanstalt, Bundessortenamt und
CHemische Industrie’’ (BBCH) extended scale. Since there were different spaces for
crop separation (17 cm, 70 cm and 75 cm for wheat, sunflower and maize,
respectively) whereas the plant morphology also varied (wheat and maize are
monocotyledonous plants, and sunflower is a dicot), the images were very different
among them, forming a complete image set to test the algorithm.
NDVI was also calculated after the application of herbicide on two plots using an
UAV in 2007 (Fig. 1)
Fig. 1. Pseudocolor NDVI map for the turf grass field on 3 dates in 2007. Source: (Xiang &
Tian, 2011).
Jannoura, R. et al., conducted a research in 2015 to evaluate crop biomass over a
field with peas and oats. An RGB digital camera was adopted on a remote -
controlled hexacopter (Fig.2) and, based on the aerial pictures, the Normalised Green
- Red Difference Index (NGRDI) was calculated and related to aboveground biomass
and Leaf Area Index (LAI). The Green - Red ration vegetation index (GRVI), the
Normalised Difference Vegetation Index (NDVI) and the Visible Vegetation Index
734
(VVI) were also calculated in the research. ( (Jannoura, R. et al., 2015)
Fig. 3. e Hexacopter used to take low-altitude aerial photographs (Jannoura, R., et al., 2015)
6 Conclusion
Results of recent studies indicate that true colour images allow determining crop
variation maps of an entire field. The results are encouraging for the development of
UAVs as a tool for site-specific precision agriculture in a small field area given their
low cost of operation. It is suggested that to provide a reliable end product to farmers
advances in platform design are required and the farmer needs to be actively involved
in image acquisition, interpretation and analysis in order to have reliable assistance in
farm management decision making.
Data analysis has to be able to explain what is causing a variation in agricultural
production, not just identify that there is a variation. The way forward for the
industry now is to ensure that we can move from UAVs simply producing data to
providing the agricultural industry with knowledge. We have to be able to produce
high precision data that can improve farming in practice if UAVs are to become a
key component of the agriculture industry. Predictions for the industry see growth
over the next 2-3 years, with UAVs fully integrated into the agriculture sector within
5 years. By 2018 agricultural UAVs are also predicted to be cheaper, autonomous
and a key part of the agriculture industry.
The future of precision agriculture is very exciting and the term Unmanned Aerial
Vehicles (UAV) will eventually be referred to as Unmanned Aerial Systems (UAS)
to correctly identify these highly engineered, safe and valuable tools that will
increase profitability for future crop production.
References
1. Christophersen, H., Pickell, R. W., Koller, A. A., Kannan, S. K., & Johnson, E.
N. (2004). Small adaptive flight control systems for UAVs using FPGA/DSP
735
technology. 3rd Unmmanned unlimited technical conference, workshop and
exhibit (pp. 1-12). Illinois: AIAA.
2. Costa, F. G., Ueyama, J., Braun, T., Pessin, G., Osorio, F. S., & Vargas, P. A.
(2012). The use of unmanned aerial vehicles and wireless sensor network in
agricultural applications. Geoscience and remote sensing symposium (pp. 5045-
5048). Munich: IEEE International.
3. Herwitz, S. R., Johnson, L. F., Dunagan, S. E., Higgins, R. G., Sullivan, D. V.,
Zheng, J., et al. (2004). Imaging from an unmanned aerial vehicle: agricultural
surveillance and decision support. Computers and Electronics in Agriculture , 44,
49-61.
4. Jannoura, R., Brinkmann, K., Uteau, D., Bruns, C., & Joergensen, R. G. (2015).
Monitoring of crop biomass using true colour aerial photographs taken from a
remote controlled hexacopter. Biosystems engineering , 129, 341-351.
5. Maurya, P. (2015, 06 01). IIT Kanpur. Hardware implementation of a flight
control system for an unmanned aerial vehicle . Kanpur, Kanpur, India. Retrieved
06 01, 2015, from Computer science and engineering:
http://www.cse.iitk.ac.in/users/moona/students/Y2258.pdf
6. Meyer, G. E., & Neto, C. J. (2008). Verification of color vegetation indices for
automated crop imaging applications. Computers and electronics in agriculture ,
63, 282-293.
7. Primicerio, J., Di Gennaro, S. F., Fiorillo, E., Genesio, L., & Lugato, E. (2012). A
flexible unmanned aerial vehicle for precision agriculture. Precision Agriculture ,
13 (4), 517-523.
8. Sugiura, R., Noguchi, N., & Ishii, K. (2005). Remote - sensing technology for
vegetation monitoring using an unmanned helicopter. Biosystems Engineering ,
90 (4), 369-379.
9. Torres-Sánchez, J., López-Granados, F., & Peña, J. M. (2015). An automatic
object-based method for optimal thresholding in UAV images: Application for
vegetation detection in herbaceous crops. Computers and Electronics in
Agriculture , 114, 43-52.
10. Wang, J., & Song, Y. (2009). Hardware design of video compression system in
the UAV based on the ARM technology. 2009 International Symposium on
Computer Network and Multimedia Technology (pp. 1-4). Wuhan: CNMT.
11. Xiang, H., & Tian, L. (2011). Development of a low-cost agricultural remote
sensing system based on an autonomous unmanned aerial vehicle (UAV).
Biosystems engineering , 108, 174-190.
736