Application of Remote Sensing in fruit

production and development

Kiran Patnaik
04 FSC/Ph.D./16

Department of FSHT , Orissa University of Agriculture and Technology, Bhubaneswar

Remote Sensing
: An eye in the
sky for
agriculture
 Agenda
 Introduction to Remote Sensing
 Brief history
 Elements of Remote Sensing
 Principle of Remote Sensing
 Components of Remote Sensing systems
 Application in fruit crops
 Role of vegetation indices in Remote Sensing
 Research findings
 Government initiatives w.r.t. Remote Sensing
 Scope of Remote Sensing
 Future thrusts
 Conclusion
 References
3
 Introduction

 Remote sensing is the collection of reliable information about

an object without being in direct physical contact with the
object.

 Remote Sensing is science of

 acquiring
 processing
 interpreting
images and related data that are obtained from ground-based,
air-or space-borne instruments that record the interaction
between matter (target) and electromagnetic radiation.

4
 Historical perspective
 1860’s - First aerial photos from balloons

 1940’s - 50’s - widespread use of aerial photos (WWII)

 1960’s - First satellite images

 1970’s - Landsat MSS, airborne scanners

 1980’s - Landsat TM,SPOT, AVHRR

 1990’s - Airborne digital cameras, RADAR, IRS C&.D, release of classified data

 2000’s - New satellites, improvements in airborne digital cameras, LIDAR

 2010’s - Higher spatial & spectral resolution, extensive use of LIDAR,

5
Hyperspectral imagery , UAV’s
 Elements involved in Remote Sensing
 Energy Source or Illumination (A)

 Radiation and the Atmosphere (B)

 Interaction with the Object (C)

 Recording of Energy by the Sensor (D)

 Transmission, Reception and Processing (E)

 Interpretation and Analysis (F)

 Application (G)
6
Principle of Remote Sensing

7
 Remote Sensing Sensors
 A sensor system records electromagnetic (EM) radiation detected
as a combination of reflected solar radiation and emitted radiation
by an object.

Sensors

Active Sensors Passive Sensors

Artificial Energy Source Natural Energy Source

e.g. Radar Systems e.g. Sensors on Satellites

SLAR, SAR Landsat, SPOT

8
Passive Sensor vs Active Sensor

9
Remote Sensing platforms

10
 Remote Sensing organisations
1. ISPRS- International Society for Photogrammetry and Remote
Sensing
2. IGARSS-International Geoscience Remote Sensing symposium
3. NASA- National Aeronautic and Space Administration (USA)
4. ESA- European Space Agency (Europe)
5. NASDA- National Space Development Agency (Japan)
6. CNES- Centre National d’stedesSpatiales (France)
7. DARA- German Space Agency (Germany)
8. CSA- Canadian Space Agency (Canada)
9. ISRO- Indian Space Research Organisation (Banglore,
Karnataka)
10.NRSA- National Remote Sensing Agency (Balanagar,
Hyderabad)
11
 Applications of Remote Sensing in fruit crops

 Assessment of crop area

 Delineation of orchards and spatial analysis
 Determination of fruit yield
 Quantification and scheduling of precise and proper fertilizer,
irrigation needs and application of pesticides
 Identification of diseases and pest attacks
 Weather study for site specific crops management in fruit crops
 Detection of growth and health of orchards on a larger scale.
 Land suitability assessment for crops
 Disaster management
 Precision horticulture
12
a) Red colour shows cultivated b) Problems inside the agricultural fields
plants on satellite image

13
 Vegetation indices
 Biophysical features of plants can be characterized spectrally by vegetation indices.

 Calculated as ratios or differences of two or more bands in the VIS, NIR and SWIR
wavelengths.

 NDVI index- used most frequently to determine the condition, developmental stages
and biomass of cultivated plants and to forecasts their yields. It has become the most
commonly used vegetation index [Wallace et al. (2004), Calvao et al.( 2004)]

 Multispectral and hyper-spectral aerial and satellite imagery helps in creating NDVI
maps, which can differentiate soil from grass or forest, detect plants under stress,
and differentiate between crops and crop stages.

14
Vegetation indices Contd.

15
Vegetation indices Contd.

16
Research findings

17
Research findings Case Study #1

Geospatial technology for acreage estimation of natural rubber and

identification of potential areas for its cultivation in Tripura state
Potential of temporal and multi-resolution satellite datasets
were explored in this project for inventory of natural rubber in
Tripura. The study revealed that natural rubber showed distinct
spectral signature on multi-date LISS-III data and it could be
delineated using temporal NDVI profile. The satellite analysis
revealed about 48,033 ha of natural rubber (more than 3 years old) in
the state as of December 2011. The mapping accuracy for the older
mature natural rubber was 96.15 per cent indicating the utility of RS
data for NR mapping and monitoring.

18
Research findings Case Study #1

Temporal LISS-III data showing Natural Rubber plantations

19
Research findings Case Study #1

NDVI profile of different vegetation classes

20
 Research findings Case Study #1

Temporal LISS-III data along with classified data showing NR Plantations 21

Research findings Case Study #2
Blueberry Orchard Delineation with High-Resolution Imagery

 Panda et al.(2016) studied the use of advanced image

processing technique and high resolution multispectral
imaging to distinguish blueberries orchards from other land
uses.
 Very high resolution 1-meter multi-spectral NAIP imagery
was used to identify blueberry orchards in three southeastern
counties (Bacon, Brantley, and Camden) in Georgia.
Advanced image processing techniques, including principal
component analysis and self-organizing map (SOM) neural
network image segmentation technique, were used in this
study.
22
Research findings Case Study #2
Blue berry orchards surrounded by forest , shrubs and grasses

23
Research findings Case Study #2

24
 Research findings Case Study #3
Remote sensing-based monitoring of chlorophyll dynamics in fruit
orchards

Pieters et al.(2017) studied the dynamics in the

canopy chlorophyll content within a pear orchard which
was monitored by creating chlorophyll maps of the orchard
from remotely sensed images at different key moments of
the growing seasons. These chlorophyll dynamics were
related to the productivity of the trees in orchard.

25
Research findings Case Study #3

26
Research findings Case Study #3

 Distribution of the canopy chlorophyll content (µg/cm2) of

productivity class 1 (left) and productivity class 4 (right) for May,
July and August. [Pieters et al.(2017)]

27
Research findings Case Study #3

RPAS images of the orchard in May (left) and in August (right), on

which the experimental trees are indicated by the circles. The outer
and inner circles representing the productivity and chlorophyll classes
repectively. [Pieters et al.,(2017)]
28
Research findings Case Study #4
Image Based Mango Fruit Detection, Yield Estimation Using
Multiple View Geometry
 Madeleine et al. (2016) investigated the use of multiple view-
point fruit detection, tracking and counting as a means to
measure and map the quantity of fruits in a mango orchard. A
total of 522 trees and 71,609 mangoes were scanned on a
Calypso mango orchard near Bundaberg, Queensland,
Australia, with 16 trees counted by hand for validation, both
on the tree and after harvest.
 The results show that single, dual and multi-view methods can
all provide precise yield estimates, but only the proposed
multi-view approach can do so without calibration, with an
error rate of only 1.36% for individual trees. 29
Research findings Case Study #4

(a) The Shrimp Robotic Platform (b) Mango Orchard Site

 The robotic platform Shrimp (a) with labels to indicate the

locations of the sensors used in this paper; a map of the mango
orchard test block (b), with yellow pins marking the 18 trees that
were selected for ground-truth field and harvest fruit counts. 30
Research findings Case Study #4
Fruit Tracking and Counting

(a) Example with Real Data (b) Close-up View 31

 Research findings
 Shrivastava et al.(2006) performed a study in Florida to delineate
citrus groves for economic assessment, using Landsat Enhanced
Thematic Mapper Plus (30 m) imagery. Their results showed a
significant correlation between citrus production/income with
remotely sensed imagery-derived citrus area coverage.

 Sharma et al.,(2007) used Indian Remote Sensing Satellite (IRS)

LISS III and IRS AWiFS (23 m and 55 m, respectively) imagery
data to develop an apple orchard database for the entire state of
Himachal Pradesh and were successful in the delineation using a
relatively low resolution imagery.

32
 Research findings
 The same low resolution IRS imageries were also used for identifying and
prioritizing the suitable sites for passion fruit plantations in the hilly districts of the
north-eastern states of India.

 O’Connel et al.,(2005) used remotely-sensed high resolution aerial photographs to

identify the tree canopy of a peach orchard for yield forecasting and the estimation
of crop water requirements. In their case, they needed to identify individual trees to
perform SSCM and hence they used high resolution images.

 Torres et al.,(2008) conducted a study to distinguish olive tree orchards using

remote sensing images by clustering assessment techniques. They used 0.25- to 1.5-
m resolution imageries to perform the advanced SSCM process for an olive orchard
that was only 2 hectares in size.

33
 Government Initiatives
Increase crop yield with use of Remote Sensing:
FASAL (Forecasting Agricultural output using Space, Agro-
meteorological and Land-based observations)

 Launched in 2006.
 Reliable crop prediction mechanism.
 A team of ISRO/Department of Space, State Remote Sensing
Applications Centers, State Agricultural universities and many
other institutions are working on FASAL.
 Remote Sensing component of FASAL- handled by the
Ahmedabad - based Space Application Centre of the Indian
Space Research Organization.
34
35
Government Initiatives Contd.
NADAMS (National Agricultural Drought Assessment and
Monitoring System)

 Provide real-time information about the prevalence and severity

of droughts at district and sub-district level in 13 key
agricultural states.

 Prepares fortnightly reports on droughts using advanced wide-

field sensors of satellites like Resourcesat-1, IRS 1C and IRS
1D.

36
Government Initiatives Contd.
NCFC (National Crop Forecast Centre)
 Launched in April 2012
 Provide estimates of crop output and assess the drought
situation in the country through latest technologies.
 Located at the New Delhi base Pusa.
 Equipped with image-processing facilities, laboratories and
software.
 Named after the well known Indian Statistician P C
Mahalanobis, founder of the Indian Statistical Institute, who
played a significant role in evolving methods to estimate crop
yields.
37
 Government Initiatives Contd.
CHAMAN (Coordinated Horticulture Assessment and
Management using Geoinformatics)
 Implemented by Mahalanobis National Crop Forecast Centre (MNCFC) in
collaboration with ISRO Centres (SAC & NRSC) and 12 state horticulture
departments, NHRDF, IMD, ICAR Centre and State Remote Sensing
Centres.
 Envisages use of satellite remote sensing data for area and production
estimation of 7 horticultural crops (Potato, Onion, Tomato, Chili, Mango,
Banana and Citrus) in 12 major states in 180 districts.
 The programme also uses GIS (Geographical Information System) tools
along with remote sensing data for generating action plans for horticultural
development (site suitability, infrastructure development, crop
intensification, orchard rejuvenation, aqua-horticulture, etc.).

38
Government Initiatives Contd.

39
Government Initiatives Contd.
Major achievements of the project include:
 Potato Production Forecast at National/State Level and Final
forecast (F2) at District Level for 5 states (Bihar, Gujarat,
Punjab, Uttar Pradesh, West Bengal)
 Mango and Citrus Orchard mapping: Using LISS III , LISS IV
and Cartosat dataset object based/Pixel-based in selected
districts of Bihar, UP & Punjab.
 Post-harvest infrastructure planning by assessing the potential
of cold storage capacity for fruits and vegetables in Bihar state.
 Aqua-horticulture: Remote sensing based assessment of
wetlands for Makhana (Foxnut) cultivation in Darbhanga
district of Bihar.
40
 Scope of Remote Sensing
 There is a great potential for airborne hyper-spectral imagery
use in estimating fruit yield and detecting plant stress as
supported by several studies [Cross et al.,(2008)].

 Polo et al. (2008) have used a low-cost tractor-mounted

scanning Light Detection and Ranging (LIDAR) system for non-
destructive recordings of tree-row structure in orchards and
vineyards to determine the tree-row volume and total surface
area (crown cover).
 The use of LIDAR imagery data for horticultural SSCM has an
immense scope because it can determine plant heights along
with spectral characteristics.
41
Scope of Remote Sensing Contd.
 According to Schupp et al. (2009), imaging technology has
the potential to cross multiple areas of tree fruit production,
such as:
Crop load assessment
Blossom or green fruit counts
Yield estimation
Determination of insect presence or disease infection and
associated eradication
Soil moisture content determination for enhanced irrigation
system design
Estimation of fertilizer, pesticide, and herbicide application
rates
Development of assisted or automated pruning and harvesting
42
strategies.
 Future Thrusts
Keeping in view the agricultural scenario in developing
countries, the requirement for a marketable RS technology for precision
agriculture is the delivery of information with the following
characteristics:
 Low turn around time (acquired, corrected and processed) ~ 24-48
hrs
 Low data cost
 High spatial resolution (at least 2m multi-spectral for 1 ha field size)
 High spectral resolution (10-20 nm for retrieving biophysical
parameters)
 High temporal resolution (at least 5-6 dates per season)
 Delivery of analytical products in simpler format 43
 Future Thrusts
 With current satellites, one can see areas that are 30 meters x 30 meters (11.1

measurements/ha), 23 x 23 meters (18 measurements/ha), 10 x 10 meters (100

measurements/ha) and 5 x 5 meters (400 measurements/ha). With future satellites,

we will be receiving data that have a variety of spatial resolutions that in some cases

will be as detailed as 1 x 1 meter or over 10000 data points per hectare.

 Integration of biophysical parameters (such as Leaf Area Index or temperature)

derived from high-resolution satellite based remote sensing data, with physical crop

growth modeling towards an operational decision support system for precision

farming.
44
 Advancements in Remote Sensing technologies
 Unmanned aerial vehicles (UAVs), commonly known as drones.

 Drones are remote controlled aircraft with no human pilot on-board.

 capable of collecting very high-resolution imagery below the cloud

level, with much more detail than the satellite imagery.

 Easy to use as the UAV essentially flies itself.

 Data processing applications are becoming less expensive and

easier to use.
45
46
Hyperspectral image from UAV

47
 FarmBeats: An IoT Platform from Microsoft for Data-
Driven Agriculture

 Farmbeats focuses on four aspects:

Farming-yield estimation
Precision irrigation
Pest-infection
Fertilizer application

 Data required to advice on those four aspects are acquired from the
farm and collected on cloud using TV White Space technology. At
the farm level a white space device such as a router is installed by
the farmer that would facilitate data collection on soil moisture,
nutrient content, etc.

48
49
 Constraints of Remote Sensing in India

 Small farm size

 Lack of success stories
 Infrastructure and institutional constraints
 Lack of technical expertise
 Heterogeneity of cropping systems
 Issues with satellite data
Not viable for small land holdings and mixed cultivation
May not be effective for pest attack information due to inherent
delay in providing data
Limitation of predictive model development
Data degradation due to atmospheric disturbances
50
 Conclusion

 Precision farming in many developing countries including

India is in its infancy but there are numerous opportunities for
adoption.

 Progressive Indian farmers, with guidance from the public

and private sectors, and agricultural associations, will adopt it
in a limited scale as the technology shows potential for raising
yields and economic returns on fields with significant
variability, and for minimizing environmental degradation.

51
References
 Baugher T, Schupp J, Travis J, Hull L, Ngugi H, Krawczyk G, Harsh M,
Reichard K, Ellis K, Remcheck J, Crassweller R, Marini R, Harper J,
Kime L, Heinemann P, Liu J, Lewis K, Hoheisel G, Jones V, Glenn
M, Miller S, Tabb A, Park J, Slaughter D, Johnson S, Landers A,
Reichard G, Singh S, Bergerman M, Kantor G and Messner W. 2009.
Speciality Crop Innovations: Progress and Future Directions; Specialty
Crop Innovations Progress Report; College of Agricultural Sciences, Penn
State University: University Park, PA,USA ,1-16.
 Calvao T and Palmeirim J M. 2004. Mapping Mediterranean scrub with satellite
imagery: biomass estimation and spectral behaviour. International Journal
of Remote Sensing, 25: 3113–3126.
 Cross J V and Walklate P J. 2008. The UK PACE scheme for adjusting the dose
to suit apple crops. Agricultural Engineering International: CIGR Journal ,
1-10.
 O’Connell M G and Goodwin I. 2005. Spatial variation of tree cover in peach
orchard. Acta Horticulturae (ISHS), 694: 203-205.
 Panda S S, Hoogenboom G and Paz J. 2016. Blueberry Orchard Delineation
with High-Resolution Imagery and Self-Organizing Map Neural Image
Classification, Agricultural Research & Technology ,3(1): 001-0010.
 Pieters C, Vanbrabant Y, Tits L and Somers B. 2017. Remote sensing based
monitoring of chlorophyll dynamics in fruit orchards:linking vitality to
productivity. Retrieved from Limo Database(http://limo.libis.be).
Contd…

 Polo J R R, Sanz R, Llorens J, Arnó J, Escolà A, Ribes-Dasi M, Masip J, Camp F, Gràcia

F, Solanelles F, Pallej T, Val L, Planas S, Gil E and Palacín J. 2008. A tractor
mountedscanning LIDAR for the non-destructive measurement of vegetative volume
and surface area of tree-row plantations: A comparison with conventional
destructivemeasurements, Precision Agriculture-Springer, 102 :128-134.
 Sharma A, Panigrahy S. 2007. Apple orchard characterization using remote sensing and GIS
in Shimla district of Himachal Pradesh. In Proceedings of Remote Sensing and
Photogrammetry Annual Conference, 11-14 September ,Newcastle, UK.
 Shrivastava R J and Gebelein J L.2006. Land cover classification and economic
assessment of citrus groves using remote sensing. ISPRS Journal of
Photogrammetry and Remote Sensing, 61: 341-353.
 Torres L G, Pena-Barragan J M, Lopez-Granados F, Jurado-Exposito M and
sFernandez-Escobar R. 2008. Automatic assessment of agro-environmental indicators
from remotely sensed images of tree orchards and its evaluation using olive
plantations. Computers and Electronics in Agriculture, 61: 179-191.
 Wallace J F, Caccetta P A and Kiiveri H T. 2004. Recent developments in analysis of spatial
and temporal data for landscape qualities and monitoring. Australian Ecology ,29:100–
107.
Thank you

54