Precision Agriculture

PS Tiwari

Introduction
Precision agriculture, sometimes called site-specific management, is an emerging technology
that allows for adjustments to address within-field variability in characteristics such as soil
fertility, soil moisture, weed intensity and insect-pest infestation. The technology has the
potential to reduce production costs through more efficient and effective application of crop
inputs. It also reduces environmental degradation by allowing farmers to apply agricultural
inputs at appropriate rates at places where these are needed. Spatial, temporal and predictive
aspects of soil and crop variability are the vital elements of precision agriculture. It involves the
sampling, mapping, analysis, and management of specific areas within a field in recognition with
spatial and temporal variability with respect to soil fertility, moisture availability, crop
characteristics and insect-pest population. Large-scale variability has long been encountered
with different cropping practices in different regions. However, precision agriculture responds to
spatial variability within individual fields or orchards. Precision agriculture practices lead to a
more cost-effective and environment friendly agriculture by:
 Increasing agricultural productivity

 Optimizing the use of restricted natural resources such as land and water and other crop
inputs such as seeds, fertilizers, herbicides and other chemicals.

 Reducing environmental degradation

 Engaging the efficient capabilities of intelligent farm machinery

 Improving the performance of farm management

Present day agriculture has two major challenges: increasing productivity to feed the growing
population and reducing the environmental degradation caused by the input-intensive
agriculture. Earth observation data has enabled achieving sustainable agriculture in the
following aspects:
 Increase area under agriculture by identifying cultivable wastelands/marginal lands
 Increase cropping intensity through improvised cropping practices like cultivation in
kharif/post-kharif fallows
 Increase productivity by providing inputs for retention/improvement of soil fertility, site-
specific management and increased area under high yielding varieties (HYVs)
 Preserve the ecosystem and eco-diversity
 Create supportive infrastructure (irrigation potential)

Precision agriculture offers the promise of increasing productivity, decreasing production cost
with optimum use of resources and minimizing the environmental impact (National Research
Council, 1997). This technology mainly benefits from the emergence and convergence of
several technologies, including the Global Positioning System (GPS), and incorporating the

1
computer software interface component of database, mapping and interpretation with
Geographic Information System (GIS) (Gibbon, 2001). Different stages of variability can be
achieved by adopting two approaches: the map-based approach and the sensor-based
approach. Sensor-based approach allows less dependency on positioning devices, an
advantage over map-based approach with higher position accuracy and real-time application
(Zhang et al., 2002).
Concepts of Precision Agriculture Systems and Requirement
Precision agriculture is based on mapping systems, on real-time sensor-actuator systems, or on
combinations of the two. Precision agriculture concepts include:
 More precise and accurate farm work by better adjustments of settings and by improved
monitoring and control mechanisms (Figure 1).

Figure 1. Development of technology level and farming strategies

 Localized application of agricultural inputs as per requirement in accordance with the
variability of soil characteristics, nutrient levels, available water, and plant growth.

 Site-specific weed and pest management.

 Automatic information acquisition and information management with well-structured

databases, geographic information systems (GIS), highly sophisticated integrated
decision-support and expert-knowledge systems connected by standardized
communication links (Figure 2).

2
 Figure 2. An integrated precision agriculture system
Precision agriculture is not a fixed system, but rather differs with different soil types, climatic
conditions, farm management systems, production levels and mechanization solutions. Before
going for precision farming approach one has to assess variability of different soil and crop
parameters. Some relevant soil and crop parameters and principles for its sensing in precision
farming are shown in Table 1.

Table 1: Relevant parameters and principles for its sensing in precision farming

Measuring
Traditional Principle for Principle for
Parameter
method contact sensing non-contact sensing
period
Soil texture Once Soil analysis Open Open
Conductivity, high
frequency, neutron
At Tensiometer, Infrared remission,
Soil moisture scattering (NIVIR - as a
cultivation gravimetric soil radar
method for scientific
studies)
At
Soil density Gravimetric Penetrometer Open
cultivation
Determination Colour measurements
Several
Nitrogen with reactants, ion-
times Spectral evaluation of plants
content of soil sensitive electrodes,
per year
solution chemosensors
Soil and plant
Other Every 5
nutrients, (3 ... 8) Open Open
analysis;
pH-value years
pH-meter

3
 Every 5 Oxidation
Carbon (Cl 3-NMR - as a method Colour analysis, VIS- and JR-
(3 ... 8) with
content for scientific studies) remission
years CO2-analysis
Height of plants and plant
Plant Conductivity (top plant to density by moving light
Inspection
growing, At soil), force measurement barners or ultra sonic
and
biomass cultivation by moved bending of measurements, radiometric
classification
development plants methods, machine vision,
spectral aerial imaging
State of Inspection (H1-, P31 and Cl3- Spectral aerial imaging,
Before
plants, plant and NIVIR - as a method for spectral remission of plants,
treatment
protection classification scientific studies) chlorophyll fluorescence
Spectral aerial imaging,
plant recognition by imaging
Inspection
Before classification by
Weed density and Open
treatment conditions (time of
classification
cultivation, position in field,
etc.)
Indirect mass
Gravimetric determination
Yield classification by
Yield Harvest mass using volume,
spectral aerial imaging
determination momentum,
radiometric methods

System Elements
Applicators (fixed rate/variable rate)
In general there are two types of machines for application of agricultural inputs in the field i.e.
fixed rate and variable rate. In fixed rate applicators, once the general recommended application
rates are fixed after calibration, the same are supposed to be applied over the whole field area.
The general recommended application rates are decided for each region on the basis of
research findings of the SAU’s/other research organizations located in those regions. On the
other hand the variable rate applicators (VRA) vary the amount of input applied on the basis of
site-specific requirement of the soil or crop. In precision farming the variable rate applicators are
mostly used to address the site-specific variability.
Many current commercial seed/fertilizer drills, sprayer controllers maintain constant/fixed
application rate by compensating for ground speed changes, a concept that was researched
more than three decades ago (Gebhardt et al., 1974; Dickey-john, 1987).
There are two approaches for variable rate applicators. One is off-line or map based approach
and the other is on-line or real-time sensor based approach.
Map-Based Systems
Map-based systems allow information to be gathered from various automatic and manual
sources. Then a desired control map can be generated to guide such field operations as
variable planting, irrigation, fertilization, or pesticide application. Map based systems use GPS
or other locator systems to establish a geographic basis (Figure 3). To use GIS or remote
sensing information, research was initiated on map-driven variable rate sprayers (Rockwell and
Ayers, 1994). Because of the resolution of the weed map, the spatial resolution of those sprayer
controllers was relatively low - the whole boom was controlled at one rate. Current post-

4
emergence sprayers have boom widths of 20 to 50 meters. The timing of weed control was
another issue. By the time the weed map was ready, field conditions may have changed.
Reports on disease control using application maps based on visual assessment of diseases
have been published by Secher (1997) and Bjerre (1999). However, if weather conditions are
favourable, diseases can quickly spread over the whole field. If critical thresholds are exceeded,
an immediate response is required, as the accuracy of sampled data sets is time sensitive. In
practice, this off-line application method causes problems, because disease maps are not
available quickly enough for timely decisions on disease control to be made. But, in online or
real-time systems for variable-rate spraying combine disease detection and application in a
single operation. Further advantages of real-time systems are that no additional costs for labour
and data management arise. However, to run such systems, an automatic disease detection
facility is necessary to provide the information about parts of the fields where diseases occur
before their incidence reaches critical thresholds.

Figure 3. Systems and components for map-based precision agriculture

Components of map-based systems include:
 Equipment position locators
 Sensors for soil, crop and weed parameter measurements
 Mapping software with color display and printing capabilities
 Controllers for map-based applications
 Actuators to perform the control operation
Positioning in Precision Agriculture
Precision agriculture requires the integration of three basic elements:
 Positioning capabilities to locate the equipment in the field
 Real-time mechanisms for controlling seed, nutrient, water, pesticide or other crop
production inputs; and
 Databases or sensors that provide information needed to develop input response to site-
specific conditions.

5
The technologies associated with first two are advanced compared with the understanding
necessary to meet the third requirement.
Positioning is the most important element in any precision agriculture system. In most cases x-y
(longitude and latitude) coordinates are sufficient. For some cases and for more sophisticated
requirements, z (elevation) may be also required. A categorization of accuracy requirements
into four different types of agricultural tasks is given in Table 2.

Table 2. Positioning accuracy requirements in different types of tasks

Required Task Examples

Accuracy

±10 m  Navigation  Targeting of fields (machinery ring, contractor)

 Targeting of storage area (forestry)

±1 m  Job  Local field operations such as yield monitoring,

execution fertilizing, plant protection, soil sampling, action in
protected areas
 Information
 Automated data acquisition
 Documenta
tion

±10 cm  Vehicle  Gap and overlap control (fertilizing, spraying)

guidance
 Harvesting without skips

±1 cm  Implement  Mechanical weed control

(tool) guidance

Satellite Navigation Systems

With the installation of satellite navigation systems by US and USSR during the late l980s,
military and civilian industries acquired access to worldwide cost-free location data available
continuously independent of daylight and weather conditions. These systems often are
generically termed as GPS after the most popular system.
Global Positioning System (GPS)
GPS is widely available in the agricultural community and its potential is growing. Farm uses
include mapping yields (GPS + combine yield monitor), variable rate planting (GPS + variable-
rate planter drive), variable rate lime and fertilizer application (GPS + variable-rate spreader
drive), variable rate pesticide application (GPS + variable-rate applicator), field mapping for
records and insurance purposes (GPS + mapping software) and parallel swathing (GPS +
navigation tool).
There are essentially three parts that make up GPS:
a) Location or space segment,
b) Management or control segment and
c) Receiver or user’s segment.
The space segment is based on the constellation of 24 active and 3 spare satellites orbiting the
earth. The control segment is a system of five monitoring stations located around the world, with
the master control facility located at Falcon Air Force Base in Colorado. The user segment,

6
which is the fastest growing segment, is made up of GPS receivers and the user community.
GPS receivers convert the satellites’ signals into position, velocity, and time. This information is
used for navigation, positioning, time dissemination, and research (Figure 4).

Figure 4. Global Positioning System (GPS)

a) Location or Space Segment
The space segment consisting of positioning satellites (currently 24) with a lifetime of up to 7
years in “blocks” of equal configuration and performance. The basic idea behind the GPS
system is to determine precise location using triangulation. Based on the intersection of a group
of satellites’ signals, triangulation or satellite ranging is used to calculate a location on earth by
measuring the distance from each of several satellites in space. The satellites act as reference
points in space. Knowing the distances from the satellites to a point on the earth’s surface
allows a position to be accurately determined. Four satellite measurements are needed to
determine exact position in three-dimensional space. In order for triangulation to work, a
receiver measures the amount of time a radio wave takes to travel from a satellite to the
receiver. Both the satellite and receiver generate a set of digital codes called pseudo-random
codes at exactly the same time. The pseudo-random code repeats itself every millisecond and
is carried on radio waves. Each satellite transmits two carrier signals. Thus, the difference
between the satellite’s code and receiver’s pseudo-random code will give the distance between
the two. Since accurate time is important to the system, each satellite carries four atomic clocks.
Only one clock on each satellite is used while the other three are backups. These clocks give
the most stable and accurate measurement of time. In addition to having accurate time, the
location of each satellite is needed in order for GPS to work. The satellites are placed high
above the earth (about 12,500 miles) so their orbits are very predictable.
b) Management or Control Segment
The control segment is one, which maintains the overall control of the system and is operated
secretly by the system owner. The ultimate accuracy of the system is determined by the sum of
several types of errors. The control segment is responsible for constantly monitoring satellite
health, signal integrity, and orbital configuration from ground. This segment includes three
sections: a) master control station, b) monitor station and c) ground antennas.
c) Receivers or User’s Segment
The GPS user’s segment consists of GPS receiver. The receiver collects and processes signals
from GPS satellites that are in view and then uses that information to determine and display the
location, speed, time and so forth. The GPS receiver does transmit any information back to the
satellite.

7
In the user segment there are two classes of receivers, military and civilian (standard). The
civilian receiver can read the L1 frequency. Military or authorized users with cryptographic
equipment, keys, and specially equipped receivers can read the L2 as well as the L1 frequency.
The combination of the two frequencies greatly increases the accuracy. Most receivers available
are continuous receivers and they can monitor four or more satellites simultaneously, depending
on the number of channels available. Continuous receivers are more expensive and require
more power, but they can give instantaneous position and velocity. There is also less possibility
of error because the receiver can receive several pseudo-random codes at the same time. The
user has to determine what level of accuracy for location and velocity is needed for the available
budget. With more channels, the receiver can gain additional accuracy but the cost is higher for
the unit.
Components and Method of Operation
The two currently available satellite navigation systems i.e. GPS (United States) and GLONASS
(Russia) have almost similar characteristics (Table 3).

Table 3. Configurations of GPS and GLONASS

GPS-NAVSTAR GLONASS

Global Positioning System— Global Navigation

Name Navigation System by Time and Satellite System
Range

Ownership USA: Department of Defense Russia: Department of Defense

Satellites 24 24

Orbits 6 3

Altitude (kin) 20,183 19,100

Coordinate system WGS84 SGS85

Approximate accuracy
for civilians (2 drms) 100a 35
(m)
a
With Selective Availability intentional degradation
Differential Corrections
The accuracy of GPS location sensing can be improved by using an additional receiver at a
fixed known position (e.g., the DGPS base station in Figure2). The additional receiver compares
the GPS indication of its position to its known location to determine the instantaneous
magnitude and direction of the GPS error. Assuming the same error is present for the receiver
at the unknown location of the moving farm equipment, the GPS indication of that receiver can
be corrected easily for the error. Different methods can be used for correction (Table 4).

Table 4. Correction methods and agricultural operations suitability

Method Mapping Operations Control Operations

Real-time

8
 Own base station Yes Yes

Commercial correction Yes Yes

Post-processing Yes No

Real-time correction signals are necessary for operations in which agricultural equipment is
controlled on the basis of position, such as fertilizer or pesticide applicator. The GPS error
correction is transmitted from the stationary receiver to the mobile (rover) receiver on the
agricultural equipment. The correction can be supplied by the rover user’s own base station,
although it should be close to the rover due to governmental restrictions on the power of private
radio links. Alternatively, a commercial or governmental correction service may be used. Less
financial investment and user maintenance is required with a correction service, although there
may be a usage charge from the correction provider.
Post processing corrects the position data after the data was collected and transferred to an
office computer. No radio link is needed, and higher accuracy might be achievable. However,
because accurate position information is not available until after the operation, control
operations cannot use this technology. In addition, corrections must be obtained and used to
process the uncorrected location data.
Non-satellite Location Systems
Other non-satellite location-sensing systems are also available which may be of regional or
farm-specific interest (Table 5).
Due to the large investment in sender installations, maintenance requirements, and the
fulfillment of legal constraints, these positioning systems are concentrated in small areas or fulfill
specific purposes in which they are competitive with satellite systems.

Table 5. Examples of non-satellite location systems

Infrared Beacon Network

Rover equipment Sender/receiver Receiver Receiver

Commercial sender
Fixed equipment Reflectors Senders
network

Range (km) <5 <600 <400

Accuracy 10-20 cm 1-3 m 20-30 cm

Approximate price
per mobile unit (US$)

Real-Time Systems
These systems do not require locators or mapping software and hardware. The relevant
parameter is sensed and then an appropriate action is immediately taken. Examples include
herbicide application based upon sensed organic matter and anhydrous ammonia application in
growing maize (corn) based upon sensed soil nitrate level. The primary limitation of real-time
system is that it uses only the current sensor data. For example, soil-type information is not

9
made available in a real time herbicide or anhydrous ammonia applicator. Prior crop-yield data
also are not made available.
A few real-time field systems have been developed. The photosensor-based plant detection
systems (Shearer and Jones, 1991; Hanks, 1996) can detect all the green plants (weed and
crop plants) and spray only on the plants. A machine-vision guided precision band sprayer for
small-plant foliar spraying (Giles and Slaughter, 1997) demonstrated a target deposition
efficiency of 2.6 to 3.6 times that of a conventional sprayer, and the non-target deposition was
reduced by 72% to 99%. For high value crops, high-accuracy machine vision and control
systems have been studied for outdoor field applications in California (Tian et. al., 1997; Lee
and Slaughter, 1998).
A system that could make use of the spatial distribution information in real-time and apply only
the necessary amounts of herbicides to the weed-infested area would be much more efficient
and minimize environmental damage. Therefore, a high spatial resolution, real-time weed
infestation detection system seems to be the solution for site-specific weed management.
In the context of machine-vision application in variable rate herbicide application Carrara et al.
(2004) designed a system which consisted of a differential global positioning system (DGPS), a
portable computer, specifically developed software and an actuation device for applying at a
rate proportionally related to the machine forward speed (Figure 5).

Figure 5. A Smart Spray Concept

A spraying system for spatially differentiated herbicide application requires a novel hydraulic
arrangement controlled by a sophisticated electronic control system (Paice et al., 1995; Stafford
and Miller, 1993). The classic technique for controlling the nozzle discharge rate is to regulate
the pressure drop across the nozzle. However, this type of pressure-based flow control system
has at least three undesirable effects, i.e., slow system response, limited flow control range, and
poor nozzle performance (Giles et al., 1996). To overcome these problems, a new technique,

10
called pulse-width modulation (PWM) flow control, was developed for flow rate control with
conventional agricultural spray nozzles (Giles and Comino, 1989; 1990). Han et al. (2000) used
the correlations; to estimate the calculated flow rate, and the relative error to conduct flow rate
calibration for solenoid operated sprayer nozzles. They developed characteristics curves of
PWM settings for nozzle operations (Figure 6).

Side View

Top View
Figure 6. Schematic of a variable rate applicator system (all dimensions in cm)
Real-Time Systems with Maps
The most sophisticated systems combine the capabilities of map-based and real-time systems.
Maps of yields, soil types, and nutrients can be used with real-time sensors of plant growth, soil
moisture, and weed infestation to control field operations. Because such systems require all the
components listed previously, they are complex and expensive. But they allow the optimization
of the field operations. Intelligent decision software that decides the proper action based upon
map and sensor data in real time is crucial.
Yield Mapping
The crop yield is the main target of production agriculture. The farmer is therefore very much
interested in knowing yields. Yield monitors are an essential component of many site-specific
management strategies for production agriculture. Yield maps, the visual and numerical product
of yield monitors, encourage creative investigations and can provide reliable answers to
carefully performed on-farm management practices. Yield maps provide feedback for
determining the effects of managed inputs such as fertilizer and lime, seed and pesticides, and
cultural practices such as tillage, irrigation and drainage.

11
A yield monitor is typically a combination of two major sub-systems, namely positioning
subsystem, which determines the position of
the combine at a particular location in the field
and yield monitoring subsystem, which
measures the yield variability with location of
the combine in the field. The two subsystems
are a combination of several components
(Figure 7). The data from these two
subsystems are merged to form a data
storage subsystem, which generates the yield
and position file for the creation of a yield
map. A user interface (display and key pad)
and a task computer located in the combine
cab, controls the integration and interaction of
these components. The sensors measure the
mass and volume of grain flow (grain flow
sensors), separator speed, ground speed, Figure 7. Components of a Yield Monitor
grain moisture and header height. The
sensors are interfaced with analog to digital and direct digital inputs. Yield is determined as a
product of the various parameters being sensed.
Grain Flow Sensors
For grain flow measurement, on-the-go sensors are available to detect either the volume flow or
the mass flow. The sensors usually are installed near the top of the clean grain elevator and are
approximated as having a 12- to 15-second time lag between when the grain is cut and the flow
sensed. Several types sensors are available (Figure 8).

Figure 8. Grain flow sensors for yield mapping

Depending on the measurement principle, the accuracy of yield sensors is influenced by factors
such as moisture, density, throughput, elevator speed, and inclination. Each sensor therefore
has its own accuracy under particular conditions (Table 6).

Table 6. Accuracy of grain yield sensors

12
 Measurement Deviations in 1 s (%)

15- 15-degree
Flat Field Tests
Sensor Type degree Up/Down
Surface (See Note)
Side Hill Hill

Yield-O-Meter Volume 1iow

1.80 1.15 1.68 3.86
(Class, D) cell wheel

CERES 2 Volume flow

0.94 3.04 9.49 3.43
(RDS, UK) light beams

Flow control Mass flow

2.24 2.10 1.10 4.07
(MF, DK) radiometric

Yield Monitor
Mass flow force 3.15 1.41 1.64 4.06
(Ag-Leader, USA)
Note: Results of tests conducted at the Technical University of Munich. Field tests from three years under
various conditions and with various combine types.
Calibration of the sensor is crucial and should be done frequently. Feedback calibration using
field-transport wagons with weighing capabilities or scale measurements will improve the
accuracy.
Other Continuous Crop Flow Sensors
Yield sensors are in various stages of development and commercialization for many crops.
 Sugar-beet and potato harvesters may use belt weighing sensors (Figure 9). Accuracy is
influenced by heavy vibrations during transport and by dirt in these root crops.

Figure 9. Sugar beet yield monitoring equipment setup and connection diagram
(based on Walter, et al.).
 Sugar-cane harvesters similarly measure the mass flowing across a conveyor portion.

13
  Forage harvesters, mainly choppers, balers, and self-loading trailers, have a throughput
of high-moisture material. High accuracy therefore also requires real-time moisture
sensing. Available systems use various sensor systems (Figure 10).

Figure 10. Forage sensing system in a chopper (left) and a round baler (right)
 Cotton yield may be measured by passing the picked boIls through a light beam, against
an instrumented plate, or by accurately measuring the change in weight of the
harvester’s storage basket.
 Other crops may be measured in similar manner, using existing sensors from the more
popular crops or by developing new sensors. Volume can be measured by having the
moving crop interrupt light beams or by a positive-displacement metering device. Mass
can be measured by the force from a momentum change, by radiation absorbance, or by
weight in a transport or storage component.
Non-Continuous Yield Mapping
Some harvesting methods (for example, hay hales or fruit bins) will result in discrete units of
yield at particular points rather than continuous crop output. If the hales or bins can be assumed
to have an equal amount of crop, their positions are simply recorded. The density of position
marks on a map indicates yield. If the assumption is invalid or better accuracy is required, the
weight of each bale or bin is recorded through the use of load cells, strain gages, or hydraulic
pressure on the loader or transport equipment.
Data Storage and Mapping
Signals from all sensors (continuous-yield, moisture, location, working-width) are processed to
give yield per area in short-period cycles (usually once per second) and stored in an on-board
controller. Signaling to the driver is necessary to monitor continued error-free functioning of the
sensors and to show the actual situation of the work.
Data transfer to the on-farm office computer can use chip cards, PCMCIA (Personal Computer
Memory Card International Association) cards, or radio links. Besides yield and positioning data,
other information about the harvested plots may be added later for farm management.
Mapping programs in the office computer then can produce tracking maps and yield maps:
Tracking maps show the work sequence and the accuracy of location sensing and can be
further analyzed for task times (Figure 11). Yield maps can be established using either grid or
contour mapping. Both type’s present similar information and the choice may depend upon the
user.

14
 Figure 11. Tracking map of a maize (corn) chopping plot (DGPS, 4.5 in working width)

Yield maps show the large-scale variations in a field (Figure 12). To understand the reality of a
certain plot, several (perhaps at least three) consecutive yield mappings are necessary (Figure
13). Strong correlations (perhaps 0.7-0.9) between different crop years confirm highly stable
yield patterns.

Figure 12. Contour yield example (16.6 ha wheat field southern Germany, 5.2 t/ha average
yield

15
 Figure 13. Consecutive relative yield maps and their pattern correlations (3.7 ha, 50 m X
50m grid size)
Soil and Weed Mapping
Information about the variability of different soil attributes within a field is essential for the
decision making process. Traditional method of collecting samples and analyzing in laboratory
is laborious and time consuming. The inability to obtain soil characteristics rapidly and
inexpensively remains one of the biggest limitations of precision agriculture. Numerous
researchers and manufacturers have attempted to develop on-the-go soil sensors to measure
mechanical, physical and chemical soil properties. The sensors have been based on electrical
and electromagnetic, optical and radiometric, mechanical, acoustic, pneumatic, and
electrochemical measurement concepts (Adamchuk, et al. 2004). While only electric and
electromagnetic sensors are widely used at this time, other technologies are also being tried to
improve the quality of soil related information in the near future.
Weed mapping for site-specific herbicide application can be part of a complete precision-
agriculture system. Weed control system integration and information flow with two application
platforms has been shown in Figure 14.

Figure 14. Weed control system integration and information flow with two application
platforms.

16
Soil Sampling
Soil sampling may be done using different strategies, such as random, aligned, zoned, and
repetitive.
 Random soil sampling involves choosing sampling points randomly within a field or a
portion of a field. Often, a field is divided into a grid and a random sample is taken within
each grid square. Random sampling minimizes some of the systematic errors possible in
sampling.
 Aligned sampling takes samples on a systematic grid in the field. Aligned sampling
minimizes the maximum distance to sample locations within a field and has simplifying
advantages during some interpolations and other analyses. But it is susceptible to
systematic errors, such as being confounded with natural or cultural patterns.
 Zoned sampling can be done after stable yield and soil patterns are established. Within
the zones of uniform yields and soils, sampling is done randomly at a lower density. This
reduces the number of needed samples and the total costs of soil sampling by assuming
that the soil properties follow the zone boundaries.
 Repetitive soil sampling tries to monitor the changes of soil nutrients over time. The
sample location is accurately determined with DGPS and revisited during subsequent
samplings in other years. The original locations of sampling may have been determined
by one of the other methods.
Weed Mapping
Weed mapping is in an earlier state of development than crop yield or soil mapping. The
following types of systems (Figure 15) can be used:

17
 Figure 15. Weed mapping systems
 Manual mapping uses the detection of weed infestations by people combined with
precise DGPS location to generate maps. Such mapping only locates areas of heavy
infestation for later pesticide application using spot sprayers.
 Plant-coverage systems detect the overall plant coverage of the soil and distinguish
between plant coverage without weeds and plant coverage with additional weeds or
between fallow fields without weeds and fallow fields with weeds. Because the sensor
measures only the amount of green plant, increased crop may be falsely sensed as
increased weeds.
 Image processing analyzes images on the basis of size, shape, color, or location and
transmits the location and type of weed or crop (Figure 16). Components of a sample
system include a CCD (charge-coupled-device) camera, high-speed image analyzer,
database with image information, controller, and sprayer actuator. Such a system is
more complex and costly but will distinguish between crop and weeds if performing
properly.
 Spectral signature based weed identification uses the spectral signatures of crop and
weeds to differentiate between crop and weed. The instruments generally used for such
systems are remote sensing equipment.

Figure 16. A schematic diagram of general image processing procedures

18
Remote Sensing
Remote Sensing (RS) refers to the science of identification of earth surface features and
estimation of their geo-biophysical properties using electromagnetic radiation as a medium of
interaction. Spectral, spatial, temporal and polarization signatures are major characteristics of
the sensor/target, which facilitate target discrimination. Earth surface data as seen by the
sensors in different wavelengths (reflected, scattered and/or emitted) is radiometrically and
geometrically corrected before extraction of spectral information. RS data, with its ability for a
synoptic view, repetitive coverage with calibrated sensors to detect changes, observations at
different resolutions, provides a better alternative for natural resources management as
compared to traditional methods. Indian Earth Observation (EO) programme has been
applications- driven and national development has been its prime motivation. From Bhaskara to
Cartosat, India’s EO capability has increased manifold. Improvements are not only in spatial,
spectral, temporal and radiometric resolutions, but also in their coverage and value-added
products. Some of the major operational application themes, in which India has extensively
used remote sensing data are agriculture, forestry, water resources, land use, urban sprawl,
geology, environment, coastal zone, marine resources, snow and glacier, disaster monitoring
and mitigation, infrastructure development, etc. Major specifications of present IRS series of
satellites are shown in Table 7.
Remote sensing uses overhead images of a field or farm to indicate weed infestations or crop
health. The images usually are acquired from satellites, airplanes, or remotely piloted vehicles.
The presence of weeds between crop plants or rows of plants may be detected, or changes in
the crop health, such as size, shape, maturity stage, or color, may be determined to indicate
variations in insects, fungal infections, salinity, drainage, or other problems. Remote sensing
can use single or multiple visible, infrared, or radar frequencies. Analyses often take the relative
reflectance of solar wavelengths for differentiation (Figure 17 & 18). Spot sprayers that turn on
and off, or vary application rate, then may be programmed to spray according to the remote-
sensing maps. Table 8 & 9 shows the Major Indian remote sensing missions for agriculture
(current & immediate future), and Important remote sensing application projects in Agriculture,
respectively.

19
 Figure 17. Spectral signature of soil (wet and dry) and wheat crop

Figure 18. Primary absorption bands and factors responsible for changes in the spectral
reflectance curve.

20
Table 7. Major specifications of present IRS series of satellites

Table 8. Major Indian remote sensing missions for agriculture (current & immediate future)
(Gowrisankar and Adiga, 2001)

Year of
Mission Sensors
launch
1988 LISS-1 (72.5 m resolution;148 km swath)
IRS-IA, lB
1991 LISS-Il (36.25mresolution;142 km swath)
IRS-P2 1994 LISS-Il (36 m resolution; 142 km swath)
PAN (5.8m resolution;70 rn swath)
1995
IRS-IC, ID LISS-TIl (23.5m, 70.5 m resolution; 141 krn, 148 km swath)
1997
WiFS (188.3 m resolution; 774 km swath)
IRS-P3 1996 WiFS (188.3 m resolution; 774 km swath)
TES 2001 PAN (1 m resolution;14 km swath)
RESOURCESAT-I 2001 LISS-IV (6 m resolution: 25 km swath)

21
 LISS-lIl /23 m resolution; 140 km swath)
AWiFS (180_m_resolution; 800_km_swath)
CARTOSAT-I 2002 PAN Stereo (2.5 m resolution;30 km swath)
CARTOSAT-2 2002/03 PAN Stereo (1 m resolution; 12 km swath)

Table 9. Important remote sensing application projects in Agriculture (Venkataratnarn, 2001)

S.N Project User organization

o
1. Crop Acreage and Production Estimation State Departments of Agriculture Ministry of
(CAPE) Agriculture
2. Cotton Acreage and Condition Assessment Ministry of Textiles.
(CACA)
3. Forecasting of Agricultural output based on State Department of Agriculture
Satellite, Agro-meteorology and Land based
observation (FASAL)
4. Nation wide mapping of soil resources on Ministry of Agriculture
1:250000 scale
5. Mapping of saline/alkaline soils of India Central and State Agricultural Departments.
6. National wasteland mapping on 1:1 Million Wasteland Development Department / District
scale/I :50,000 scale Authorities
7. Nationwide mapping of land use/land cover Planning commission
for agro-climatic zones on 1:2,50,000 scale
8. National Agricultural Drought Assessment Department of Agriculture and
and Management Systems (NADAMS) Co-operation, State Agricultural
Departments and District Authorities.
9. Forest cover mapping for the entire country Department of Forest, Forest Survey of India,
State Forest Departments.
10. National drinking water technology mission on Ministry of Rural Development, Ground Water
1:2,50,000 scale Department
11. Satellite monitoring of command areas for Central Water Commission / State
crops, salinity/alkalinity and waterlogged Command Area Development Authorities.
areas
12. Integrated Mission for Sustainable Planning Commission, Ministries of Rural
Development (lMSD) Development, Agriculture, Environment and
Forests, State governments.

Remote Sensing Techniques and Capabilities

Two types of remote sensing technologies have been used in acquiring spectral images. One
type uses broadband spectral imaging from aircraft or satellites. This approach has been
successfully used for years. But its sensitivity is limited by the relatively low spatial and spectral
resolution of the satellite images (Moran et al., 1997). This technique is suitable to monitor
forest damage and crops grown in large areas. The other group is ground based, which use
narrow spectral imaging bands with imaging cameras located just a few meters above the crop
canopy (Carter and Miller, 1994). Higher spatial and spectral resolution can be achieved with
this approach. Recent literature has shown that the narrow bands may be crucial for providing
additional information with significant improvements over broad bands in quantifying agricultural
crops.

22
Remotely sensed data can be of many forms, including variations in force distribution, acoustic
wave distribution or electromagnetic energy distributions and can be obtained from a variety of
platforms, including satellite, airplanes, remotely pilot vehicles, handheld radiometers or even
bucket trucks. They may be gathered by different devices, including sensors, film camera, digital
cameras, and video recorders. Our eyes acquire data on variations in electromagnetic
radiations. Sensors can be differentiated in: Passive sensors: without their own source of
radiation. They are sensitive only to radiation from a natural origin. Active sensors: which have a
built in source of radiation. Examples are Radar (Radio detection and ranging) and Lidar (Light
detection and ranging). This can be analogue (photography) or digital (multispectral scanning,
thermography, radar). The elements of a digital image are called resolution cells (during the
data acquisition) or pixels (after the image creation). The implementation of remote sensing data
by the user requires some knowledge about the technical capabilities of the various sensor
systems. The technical capabilities of the sensor systems can be listed in three resolutions:
 Spatial resolution: concerns the size of the resolution cell on the ground in the direction
of the flight and across. The size of the pixel determines the smallest detectable terrain
feature.
 Spectral resolution: concerns the number, location in the electromagnetic spectrum
and bandwidth of the specific wavelength bands or spectral bands. This resolution differs
from sensor to sensor and largely determines the potential use of the sensor system.
 Temporal resolution: concerns the time lapse between two successive images of the
same area. This primarily determined by the platform used, and secondly by the
atmospheric conditions.
Spectral Signature
Spectral Signatures are the specific combination of reflected and absorbed electromagnetic
(EM) radiation at varying wavelengths which can uniquely identify an object. The spectral
signature of an object is a function of the incidental EM wavelength and material interaction with
that section of the electromagnetic spectrum. The measurements can be made with various
instruments, including a task specific spectrometer, although the most common method is
separation of the Red, Green, Blue and Near Infrared portion of the EM spectrum as acquired
by digital cameras. Calibration of spectral signatures under specific illumination are collected in
order to apply an empirical correction to airborne or satellite imagery digital images.
When using a spectrometer, one will look through it at a tube of gas with high voltages passing
through it. In the Spectrometer, one will be able to see specific rays of colour falling on
graduations on the inside. Each substance will have its own unique falling of rays.
Most remote sensing applications process digital images to extract spectral signature at each
pixel and use such signature to divide the image in groups of similar pixels (segmentation) using
different approaches, and, as last step, assign a class to each group (classification) by
comparing with known spectral signatures. Depending on pixel resolution, a pixel can represent
many spectral signatures "mixed" together - that is why a lot of remote sensing analysis is done
to "unmix mixtures". Ultimately correct matching of spectral signature recorded by image pixel
with spectral signature of existing elements leads to accurate classification in remote sensing.
The potential for using plant reflectance spectra in precision agriculture has been the focus of
many researchers for some time. Applications that have been investigated include nutrient
management, monitoring yield quantity and quality, and the detection of weeds for selective
herbicide application.
A suitable multi-spectral imaging system is critical in order to apply spectroscopy technique at
the plant canopy level. Recently, due to the advancement in optics and computer

23
miniaturization, researchers have integrated both satellite-based and groundbased
technologies. This combines the advantages of both narrow band and broad band spectroscopy
techniques and encouraging results have been obtained. Evans et al. (1998) developed a
hyper-spectral imaging system based on a liquid crystal tunable filter (LCTF). This system,
which is used in this study, uses LCTF to achieve variable and narrow band filtering of the
reflected light. Images are taken with video cameras located a few meters from the plants,
achieving maximum spatial and spectral resolution. Several applications of this system for
evaluating plant biochemical data and stress status have been reported (Thai et al., 1998,
2000).
The factors that can be estimated by analyzing the spectral response of plants and are of major
importance for crop management were:
- biomass
- canopy water content
- carbon / nitrogen status
- other major biochemical constituents (Mg, Na, Ca, P)

For any given material, the amount of solar radiation that is reflected (absorbed, transmitted) will
vary with wavelength. This important property of matter allows us to separate distinct cover
types based on their response values for a given wavelength. When we plot the response
characteristics of a certain cover type against wavelength, we define what is termed the spectral
signature of that cover. The various Vegetation Indices used for crop weed discrimination have
been given in Table 10. The figure below illustrates the spectral signatures for some common
cover types (Figure 19).

Table 10. Vegetation Indices used for crop weed discrimination.

Indices Formula References

Brightness Index (BI) BI = G+R+NIR+SWIR Zhengwei et al., 2008
Red Green Ratio Index (RGRI) RGRI = R/G -
Green Ratio Vegetation Index GRVI = NIR/G Oberti and
(GRVI) Baerdemaeker, 2000
Red Ratio Vegetation Index RRVI = NIR/R Oberti and
(RRVI) Baerdemaeker, 2000
Modified Green Ratio Vegetation MGRVI = SWIR/G -
Index (MGRVI)
Ratio Drought Index (RDI) RDI = SWIR/NIR Hunt and Rock, 1989
Normalized Difference Vegetation NDVI = (NIR-R)/(NIR+R) Stafford and Bolam,
Index (NDVI) 1998
Weighted Vegetation Index (WVI) WVI = NIRcrop – Stafford and Bolam,
[(NIRsoil/Rsoil)*Rcrop] 1998
Green Normalized Difference GNDVI = (NIR-G)/(NIR+G) Gitelson et al., 1996
Vegetation Index (GNDVI)
Modified Green Normalized MGNDVI = Gitelson et al., 1996
Difference Vegetation Index (SWIR-G)/(SWIR+G)
(MGNDVI)
Ratio Normalized Difference RNDVI = (NIR2-R)/(NIR+R2) Gong et al., 2003
Vegetation Index (RNDVI)
Modified Normalized Difference MNDVI = Rouse et al., 1973

24
Vegetation Index (MNDVI) (SWIR-R)/(SWIR+R)
Ratio Vegetation Index (RVI) RVI = NIR/R Jordan, 1969
Modified Ratio Vegetation Index MRVI = SWIR/R -
(MRVI)
Normalized Difference Red NDRGI = (R-G)/(R+G) Zhengwei et al., 2008
Green Index (NDRGI)
Normalized Difference Vegetation NDVSI = [NIR-(R+G)*0.5]/ Zhengwei et al., 2008
Structure Index (NDVSI) [NIR+(R+G)*0.5]
Transformed NDVI (TNDVI) TNDVI = [(NIR-R)/(NIR+R)] Tucker, 1980
Optimal Soil Adjusted Vegetation OSAVI = Rondeaux et al., 1996
Index (OSAVI) (NIR-R)/(NIR+R+0.16)
Soil Adjusted Vegetation Index SAVI = Bennedsen and Guiot,
(SAVI) [(NIR-R)*1.5]/(NIR+R+0.5) 2001
Specific Leaf Area Vegetation SLAVI = NIR/(R+SWIR) Lymberner et al., 2000
Index (SLAVI)
Modified Photochemical MPRI = (G-R)/(G+R) Zhengwei et al., 2008
Reflectance Index (MPRI)
Normalized Difference Moisture NDMI = Gao, 1996
Index (NDMI) (IR-SWIR)/(IR+SWIR)
2-band Enhanced Vegetation EVI2 = 2.5(IR-R)/(IR+R+1) Jiang et al., 2007
Index (EVI2)
Vegetation Fraction & Density VFDEI = (G-R)/(G+R-B) Gitelson et al., 2001
Estimation Index (VFDEI)
Note: NIR – Near Infrared, SWIR – Short Wave Infrared, IR – Infrared, R- Red, G – Green, B -
Blue

Figure 19. Generalized spectral signatures for some common cover types

25
By comparing the response patterns of different features we may be able to distinguish between
them, where we might not be able to, if we only compared them at one wavelength. For
example, water and vegetation may reflect somewhat similarly in the visible wavelengths but are
almost always separable in the infrared. Spectral response can be quite variable, even for the
same target type, and can also vary with time (e.g. "green-ness" of leaves) and location.
Knowing where to "look" spectrally and understanding the factors which influence the spectral
response of the features of interest are critical to correctly interpreting the interaction of
electromagnetic radiation with the surface.

With knowledge of the spectral reflectance characteristics of the earth's cover types, we can
identify and map them in areas we are generally unfamiliar with for example water, vegetation
and soil.

- Water - Longer wavelength visible and near infrared radiation is absorbed more by water
than shorter visible wavelengths. Thus water typically looks blue or blue-green due to
stronger reflectance at these shorter wavelengths, and darker if viewed at red or near
infrared wavelengths. If there is suspended sediment present in the upper layers of the
water body, then this will allow better reflectivity and a brighter appearance of the water.
- Vegetation - A chemical compound in leaves called chlorophyll strongly absorbs
radiation in the red and blue wavelengths but reflects green wavelengths. Leaves appear
"greenest" to us in the summer, when chlorophyll content is at its maximum. The internal
structure of healthy leaves act as excellent diffuse reflectors of near-infrared
wavelengths. If our eyes were sensitive to near-infrared, trees would appear extremely
bright to us at these wavelengths.

- Soil - Soils tend to have reflection properties that increase approximately monotonically
with wavelength. They tend to have high reflectance in all bands. This of course is
dependent on factors such as the colour, constituents and especially the moisture
content. As soils moisture content increases, the overall reflectance of that soil tends to
decrease. Soils rich in iron oxide reflect proportionally more of the red than other visible
wavelengths and therefore appear red (rust colour) to the human eye. A sandy soil on
the other hand tends to appear bright white in imagery because visible wavelengths are
more or less equally reflected, when slightly less blue wavelengths are reflected this
results in a yellow colour.

Control of Field Operations

Yield, crop, soil, or pest-mapping indicates non uniformities. The farmer may use those maps to
develop understanding or make decisions. Often those decisions will be to perform field
operations in a manner varied to correspond to those no uniformities. Controlling fertilization,
pesticide application, tillage, planting, or irrigation is a common part of precision agriculture.
Fertilizer and pesticide application have seen the most common implementations of precision
control, because suitable equipment is available and nutrients and chemicals should only be
applied as needed for both economic and environmental reasons.
Requirements and System Components
Field operation control requires an accurate real-time locator to indicate the equipment’s
position in the field, a map of the designed operation set points in a control computer, and an
actuator on the equipment that can implement the controller’s commands. The location must be
accurate and the dynamics of the equipment must be compensated for so that the exact action
occurs where it is desired.

26
Fertilization
The application rate of granular fertilizers (and similar materials such as lime) usually is varied
by a microprocessor controlling hydraulics that actuates a variable-speed metering wheel, a
variable-position gate, or a variable-speed chain conveyor. If the mixture of the nutrients is to be
varied as well as the rate, multiple product bins and delivery systems are needed on the
applicator, and the material-handling system must be able to efficiently refill the multiple bins.
Liquid fertilizer may be applied in a similarly variable manner. Variable rates can be achieved by
varying either pump speeds, recirculation valve flow, or flow at individual nozzles through a
variable pressure drop before the nozzle or turning the nozzle flow on and off with pulse-width
modulation.
Both granular and liquid applicators must be designed for adequate dynamic response and to
make sure the spread pattern is satisfactory at varying application rates. Variable mixture
applicators can be designed to mix near nutrient storage, although then there will be significant
time delays for changed mixtures to be transmitted to the soil, or mix near the nozzles or other
distribution devices, although then care must be taken to get complete mixture and additional
hoses or conveyors are required (Figure 20).

Figure 20. Illustration of the structure and signal flow of the control system for network-
based variable-rate fertilizer applicator.

Pesticide Application
Liquid pesticides may be applied either by varying the amount of premixed pesticide carrier
liquid or by injecting a variable amount of pesticide into a relatively constant flow of carrier
(Figure 21). Many of the same design concerns for fertilizer applicators apply to pesticide
applicators, whether liquid or granular.

27
 Figure 21. Sprayer control system for site-specific herbicide application.

Other Controller Operations

Planting or seeding also may be controlled variably to achieve precision agriculture goals by
varying variety, often in response to soil type or topographic position; population in response to
varying productivity potential or depth to find sufficient moisture.
Depth would be determined by real-time sensing while variety and population control would be
more likely to be map-based.
Irrigation may be variably controlled by either maintaining full flow but changing the length of
time of application, or maintaining length of time of application while reducing water flow rate.
Tillage and land forming operations also may be controlled in either real-time or map-based
systems.
Information Management
Sophisticated precision farming integrates a variety of computerized tools. Safe and reliable
information transfer among all these tools needs standardized communication lines
standardized interfaces, and powerful software tools.

Bus Systems on Mobile Equipment

The Advancements in agricultural electronics have led to a wide variety of controllers and
electronic components. For example, fast and reliable communications are required between a
tractor and the various implements attached to it (Figure 22).
Compatibility is insured by communications standards. Two such standards, both Network
(CAN), are as follows:

28
Compatibility is insured by communications standards. Two such standards, both using
Controller Area Network (CAN), are as follows.
The Gernman LBS (Landwirtschaftliches BUS-System [Agricultural BUS-System]), codified as
DIN 9684/2-5, is based on the 11-bit identifier of CAN V2.0A. It connects a maximum of 16
controllers, including the user terminal.
The ISO 11783 standard works with the extended identifier of CAN V2.OB and is able to
connect a maximum of 32 controllers. Its detailed structure using the ISO/OSI layer model and
an additional six parts of special definitions tries to cover all requests of agricultural tractor—
implement combinations.
Compatibility between both standards is achieved in the overall function and in the physical
layer.

Figure 22. Communication between tractor and implement using DIN 86924/2-4

Data Transfer to and from Farm Management

Complete precision agriculture systems are centered in the office computer. All information goes
to and from this main unit. Yield and soil maps are displayed on it and desired control maps are
generated on it. Data transfer can be done using human media transfer with chipcards or
PCMCIA cards, or by bringing a portable computer to the office computer. Alternatively, radio
links can be used. OPS differential corrections, remote sensing data, soil test laboratory results,
and so forth can be obtained from e-mail or the Internet.
Data Management and Geographic Information Systems
The capabilities of computerized data gathering generate large volumes of data, which must he
handled efficiently. Because precision agriculture data has position attributes, it usually is
manipulated by Geographic Information System (GIS). Geographic information systems is
defined as an information system that is used to input, store, retrieve, manipulate, analyze, and
output geographical referenced data or geospatial data, in order to support decision making for
planning, management and use of natural resources, environment, transportation, urban
facilities and other administrative records.
Such a system’s representation of a field may contain different layers of soil type and
topography, pH and cation exchange capacity (CEC), crop yields, weed maps, fertilizer and
pesticide application maps (Figure 23).

29
 Figure 23. System architecture of weed manager and links to the data sharing
environment.

These various layers can be analyzed or combined manually or automatically to generate a

control map for field operations (Figure 24).

Figure 24. Layers for the generation of application maps.

30
Decision-support Systems
Decision support systems (DSS) are a specific class of computerized information system that
supports business and organizational decision making activities. The control maps for map-
based field operations must be generated according to some sort of decision making system.
Even if the decisions are made manually, the volumes of data and the complexities of crop
production favor a decision-support system. For example, the phosphorous application rates
were calculated by a computer program for each area to remedy deficiencies and to provide
sufficient nutrients for a crop of wheat. The input data included soil type, soil-test data, and yield
potential based upon past yields.
The decision-making computer program can be deterministic, based upon rules or formulas.
The computer determines the correct control action for each small part of the field or orchard
based upon the geographic information system’s data layers and the guidelines written into the
decision making program. It also can be stochastic, based upon computer simulations.
Validated crop-growth models are run with different field-operation control strategies for
representative weather scenarios in each field portion. The strategy with the maximum
economic return and acceptable risk is used to establish the field operation control map.
Real-time systems must have control algorithms that immediately vary the actuator to the
appropriate output based upon the sensor data.

References
Adamchuk, V. I., Hummel, J. W., Morgan, M. T. and Upadhyaya, S. K., 2004. On-the-go soil
sensors for precision agriculture. Computers and Electronics in Agriculture 44: 71-91.
Auernhammer, H., ed. 1994. GPS in agriculture [special issue]. Computers and Electronics in
Agriculture 11(1).
Auernhammer, H. and Frisch, J., ed, 1993: Landwirtschafthiches BUS-SysteqLBS [Agricultural
BUS-System LBSJ. Arbeitspapier: Muenster-Hiltrup.
Basso, B., Cammarano, D., and Vita, P.D., 2004. Remotely sensed vegetation indices: theory
and applications for crop management. Rivista Italiana di Agrometeorologia 36-53 (1).
Bjerre, K. D., 1999. Disease maps and site-specific fungicide application in winter wheat. In J. V.
Stafford (Ed.), Precision Agriculture ‘99: Proceedings of the 2nd European Conference
on Precision Agriculture (pp. 495–504). Sheffield, UK: Sheffield Academic.
Carrara, M., Comparetti, A., Febo, P., and Orlando, S., 2004. Spatially variable rate herbicide
application on durum wheat in Sicily. Biosystem Engineering, 87(4): 387-392.
Carter, Gregory A. and Miller, L. R, 1994. Early detection of plant stress by digital imaging within
narrow stress-sensitive wavebands. Remote Sensing Environment, 50:295-302.
Dammer, K.H, Thole, E.H., Volk, E.T. and Hau, E.B., 2009. Variable-rate fungicide spraying in
real time by combining a plant cover sensor and a decision support system. Precision
Agric (2009) 10:431–442
Dickey-john Corp. 1987. DjCCS 100 Liquid Sprayer Control System Installation and Operation
Manual. Auburn IL: Dicky-john Corp.

31
Evans, M. D., Thai, N. C. and Jason, C. G., 1998. Computer control and calibration of a liquid
crystal tunable filter. ASAE Paper No. 973142, ASAE, St. Joseph, MI.
Gebhardt, M.R., C.L. Day, C.E. Goering and L.E. Bode. 1974. Automatic spray control system.
Transactions of the ASAE 17(6):1043-1047.
Gerhards, R., Sokefeld, M., Timmermann, C., Kuhbauch, W. and Williams II, M.M., 2002. Site
specific weed control in maize, sugar beet, winter wheat and winter barley. Precision
Agriculture, 3, 25-35.
Gibbon, D.L., 2001. Editorial: Precision agriculture and information technology. Computers and
Electronics in Agriculture, 30: 5-7.
Giles, D. K. and Slaughter, D. C., 1997. Precision band sprayer with machine-vision guidance
and adjustable yaw nozzles. Transactions of the ASAE 40(1):29-36.
Giles, D.K., Henderson, G.W., and Funk, K., 1996. Digital control of flow rate and spray droplet
size from agricultural nozzles for precision chemical application. In: Proc. Of the 3rd
International Conference on Precision Agriculture. ASA Misc. Pub1. ASA, CSSA, and
SSSA, 729-738. Madison, WI.
Giles, D. K., and Comino, J. A., 1990. Droplet size and spray pattern characteristics of an
electronic flow controller for spray nozzles. Journal of Agricultural Engineering Research
47: 249-267.
Giles, D.K., and Comino, J. A., 1989. Variable flow control for pressure atomization nozzles.
Transactions of SAE, Journal of Commercial Vehicles, 98(2): 237-249.
Griepentrog, H.W., 2005. Weed control by autonomous vehicle systems. In Proceedings:
Workshop Precision Farming, pp. 6-9.
Grisso, R. B., Oderwald, R., Alley, M., and Heatwole, C., 2003. Precision Farming Tools: Global
Positioning System (GPS). ID 442-503.
Gowrisankar, D. and Adiga, S., 2001. Remotesensisng in agricultural applications: An overview.
Proc-The first National Conference ob Agro-Informatics (NCAI), INSAIT, Dharwad, pp 9-
14.
Han, S., Hendrickson, L., and Ni, B., 2000. A variable rate application system for sprayers. In
Proc. of the fifth International Conference on Precision Agriculture. ASA Misc. Publ.,
ASA, CSSA, and SSSA, Madison, WI.
Hanks, J.E., 1996. Smart sprayer selects weeds for elimination. Agricultural Research. 44(4):15.
Hellebrand, H.J. Sensing principles for site-specific data acquisition. Institute of Agricultural
Engineering Bornim (ATB), Potsdam.
Kleinhenz, B., Jo¨rg, E., Kluge, E., and Rossberg, D., 1995. PASO—Rechnergestu¨tzte
Entscheidungshilfen fu¨ r den Pflanzenschutz. Gesunde Pflanzen, 47, 222–230.
Lee, W. and D. Slaughter. 1998. Plant recognition using hardware-based neural network. ASAE
paper No. 983040.

32
Moran, M. S., Inoue, Y., and Barnes, E. M,. 1997. Opportunities and limitations for imagebased
remote sensing in precision crop management. Remote Sensing Environment. 61: 319-
346.
National Research Council (NRC), 1997. Precision Agriculture in the 21st Century. Committee
on Assessing Crop Yield: Site-specific Farming, Information System and Research
Opportunities, National Academy Press, Washington DC., USA
Navalgund, R.R., Jayaraman, V. and Roy, P.S., 2007. Remote sensing applications: An
overview. Current Science, Vol. 93, No. 12.
Newe, M., Meier, H., Johnen, A., and Volk, T., 2003. proPlant expert.com: The online
consultation system on crop protection in cereals, rapeseed, potatoes and sugar beet: A
concept that meets the requirements of farmers and consultants in both Germany and
Europe. Bulletin OECP/EPPO Bulletin, 33, 443–449.
Nobel, S.D., Brown, R.B. and Crowe, T.G., 2002. The use of spectral properties for weed
detection and identification – A review. AIC 2002. Paper No. 02-208
Paice, M.E.R., Miller, P. C. H, and Bodle, J. D., 1995. An experimental sprayer for the spatially
selective application of herbicides. Journal of Agriculture, 60: 107-116.
Parsons, D.J., Benjamin, L.R., Clarke, J., Ginsburg, D., Mayes, A., Milne, A.E. and Wilkinson,
D.J., 2009. Weed Manager—A model-based decision support system for weed
management in arable crops. Computers and electronics in agriculture 65: 155–167.
Patil, V.C., Maru, A., Shashidhara, G.B., and Shanwad, U.K. Remote Sensing, Geographical
Information System and Precision Farming in India: Opportunities and challenges.
Department of Agronomy, University of Agricultural Sciences, Dharwad.
Robert, P. C., Rust, R. H., and Larson. W. B., ed. 1993: Proceedings qftfle jst Workshop on Soil
Specific Crop Management. Madison, WI: ASAJCSSA/SSSA.
Rockwell, A.D. and P.D. Ayers. 1994. Variable rate sprayer development and evaluation.
Applied Engineering in Agriculture. 10(3):327-333.
Rouse, J.W., Haas, R.H., Schell, J.A. and Deering, D.W., 1973. Monitoring vegetation systems
in the great plains with ERTS. In ‘Proceedings of the 3rd ERTS symposium’. NASA SP-
351 1, pp. 309–317.
Schueller, J. K. 1992. A review and integrating analysis of spatially-variable eon1p • of crop
production. Fertilizer Research 33:1—34.
Scotford, I.M. and Miller, P.C.H., 2005. Applications of spectral reflectance techniques in
Northern European Cereal Production: A Review. Biosystem Engineering, 90 (3), 235-
250.
Searcy, S. W., ed. 1991. Automated Agriculture for the 21st Century. ASAB Pukli4 lion 11—91.
St. Joseph, MI: ASAB.
Secher, B. J. M., 1997. Site specific control of diseases in winter wheat. Aspects of Applied
Biology, 48, 57–65.

33
Shearer, S. A. and Jones, P. T., 1991. Selective application of post-emergence herbicides using
photoelectrics. Transactions of the ASAE 34(4):1661-1666.
Shibusawa, S. Precision farming approaches to small-farm agriculture. Faculty of agriculture,
Tokyo University of agriculture and technology 3-5-8 saiwai- cho, fuchu, Tokyo183-8509,
Japan.
Singh, K., Agrawal, K.N. and Bora, G.C., 2011. Advanced techniquies for weed and crop
idendification for site specific weed management. Bio systems Engineering 109, 52-64.
Stafford, J. V., ed. 1996. Spatially variable field operations [special issue], Campi4crj and
Electronics in Agriculture 14(2/3).
Stafford, J.V., and Miller, P. C. H. 1993. Spatially selective application of herbicide to cereal
crops. Computers and Electronics in Agriculture, 11: 23-36.
Tangwongkit, R., Salokhe, V. and Jayasuriya, H. “Development of a Tractor Mounted Real-time,
Variable Rate Herbicide Applicator for Sugarcane Planting”. Agricultural Engineering
International: the CIGR Ejournal. Manuscript PM 06 009. Vol. VIII. June, 2006.
Thai, Chi N, 2000. Influence of canopy types on spectral evaluation of stress in plants. ASAE
Paper No. 003020, ASAE, Milwaukee, WI.
Thai, Chi N., M. D. Evans and Xiaode Deng, 1998. Visible and NIR imaging of bush beans
grown under different nitrogen treatments. ASAE Paper No. 983074, ASAE, St. Joseph,
MI.
Tian, L. Sensor-Based Precision Herbicide Application System. University of Illinois, 1304 W.
Penn. Ave., Urbana, IL 61801, USA.
Tian, L., D. C. Slaughter and R. F. Norris. 1997. Outdoor field machine vision identification of
tomato seedlings for automated weed control. Transactions of the ASAE 40(6): 1761-
1768.
Venkataratnam, L. 2001. Remote sensing and GIS in Agricultural Resources Management.
Proc- The first National Conference ob Agro-Informatics (NCAI), INSAIT, Dharwad, pp
20-29.
Verreet, J. A., Klink, H., and Hoffmann, G. M., 2000. Regional monitoring for disease prediction
and optimisation of plant protection measures: The IPM wheat model. Plant Disease, 8,
816–826.
Yaha, A., Zohadie, M., Kheiralla, A.F., Giew, S.K. and Boon, N.E., 2009. Mapping system for
tractor-implementation performance. Computers and Electronics in Agriculture.
Yuan, J., Liu, C., Li, Y., Zeng, Q. and Zha, X.F., 2010. Gaussian processes based bivariate
control parameters optimization of variable-rate granular fertilizer applicator. Computers
and Electronics in Agriculture 70, 33-41.
Zhang, N., Wang, N., Wei, J., and Stoll, Q., 2002. A real-time, embedded, weed-detection and
spray control system. Computers and Electronics in Agriculture, 36: 113-132.

34