Table Of Contents

[Link]. TITLE PAGE

NO.
1 1. Introduction 2
1.1 Motivation 3
1.2 Objective 4
2 2. Literature Review 5
2.1 Challenges in Agricultural Monitoring and Land 5
Assessment
2.2 Advancements in Remote Sensing and Geospatial 6
Analytics
2.3 Review of Existing Geospatial Platforms and 7
Index Calculators
3 System Overview and Design 8
3.1System Architecture 8
3.1.1 Overall Architecture of SoilSense 8
3.1.2 Frontend ([Link]) Interface Layer 9
3.1.3 Backend ([Link] + Python) Processing Layer 10
3.1.4 Google Earth Engine (Gee) Integration Layer 10
3.1.5 Database Layer (MongoDB) 10
3.2 Functional Workflow of SoilSense 11
3.2.1 Area Selection and Index Computation via 11
GEE
3.2.2 Data Retrieval and Index Computation via 12
GEE
3.2.3 Result Visualization and Interpretation 13
3.2.4 User Authentication and Calculation History 16
3.2.5 Data Persistence and Retrieval from Database 17
3.3 Features and Outputs 18
3.3.1 Interactive User Interface 18
3.3.2 Vegetation and other Index Trends Over Time 19
3.4 Individual Contributions 20
4 Conclusion 22
5 Reference 23
6 Biodata 25
7 Plagiarism Report 28

1|Page
1. Introduction
Agriculture is a cornerstone of global sustenance, playing a critical role in ensuring food
security for a growing population. With increasing pressure on agricultural systems due to
climate change, resource limitations, and the expanding global population, sustainable
and data-driven approaches to farming are more important than ever. One key aspect of
sustainable agriculture is understanding and managing land and soil health—factors that
are essential for crop growth, environmental stability, and long-term productivity.

However, traditional methods for assessing soil and vegetation health are often
inaccessible, expensive, and slow, especially for farmers in rural or underdeveloped
regions. These conventional methods typically rely on laboratory soil testing or manual
field surveys, which can be logistically challenging and provide only localized insights
with limited temporal relevance. This creates a knowledge gap that can result in poor
decision-making, inefficient resource use, and declining land productivity.

SoilSense is designed to bridge this gap using modern geospatial and cloud technologies.
It is a web-based platform that leverages the power of Google Earth Engine (GEE) and
satellite data to provide real-time, region-specific insights into soil and vegetation indices
such as NDVI, NDMI, NDSI, GCI, EVI and AWEI. By allowing users to select a specific
area on an interactive map and define a time range, SoilSense processes satellite imagery
and computes meaningful environmental indices to help users monitor land health over
time.

The platform is aimed not just at individual farmers, but also at agricultural researchers,
policy planners, and sustainability advocates who need efficient tools for land monitoring.
With an intuitive frontend built using [Link] and a robust backend powered by [Link],
Python, and MongoDB, SoilSense ensures scalability, responsiveness, and data accuracy.
It also integrates user authentication and a dashboard for managing historical analysis,
making it suitable for long-term monitoring and research.

The overarching goal of SoilSense is to democratize access to remote sensing analytics—

offering a cost-effective, user-friendly solution that helps users make informed decisions
based on accurate, up-to-date geospatial data. By combining satellite intelligence with
cloud computing and modern UI design, SoilSense aims to promote smarter, more
sustainable agricultural practices while reducing barriers to data-driven land management.

2|Page
1.1Motivation
Agriculture supports the livelihood of nearly 58% of the Indian population and
contributes significantly to the nation’s GDP[1]. Despite its critical importance, Indian
agriculture faces numerous challenges, including inefficient land use, overexploitation of
resources, and the inability to accurately assess soil and vegetation health. A typical
smallholder farmer owns 1-2 hectares of land and relies heavily on traditional methods,
leading to limited productivity and unsustainable practices.
One of the major obstacles is the lack of accessible tools for assessing land fertility and
vegetation health, which hinders optimal crop rotation and effective land management.
Overuse of fertilizers and water resources has led to soil degradation, reduced
agricultural productivity, and escalating costs for farmers. Furthermore, the inability to
monitor vegetation health and detect early signs of pest infestations or crop stress results
in significant yield losses and financial strain.
In this context, the use of modern technology, such as satellite-based remote sensing and
geospatial analysis, presents a transformative opportunity. By leveraging indices like the
Normalized Difference Vegetation Index (NDVI), farmers and policymakers can gain
accurate and timely insights into land and vegetation health. This data-driven approach
can enable precision agriculture, sustainable resource management, and improved crop
yields, contributing to a more resilient agricultural sector and enhancing the livelihoods
of millions of farmers.

3|Page
1.2 Objective
The primary objective of this project is to leverage the Normalized Difference
Vegetation Index (NDVI) and the computational capabilities of Google Earth Engine
(GEE) to analyze and assess ground fertility and vegetation health. This will be
achieved through the following specific objectives:
1. Develop an NDVI-Based Analysis Framework:
Using satellite imagery to calculate NDVI values and generate vegetation health
maps for the target regions.

2. Assess Soil Fertility Patterns:

Identify spatial variations in soil fertility by correlating NDVI values with vegetation
density and health.

3. Monitor Agricultural Productivity:

Track temporal changes in vegetation cover to evaluate agricultural performance and
identify areas of concern, such as degraded lands or underperforming fields.

4. Promote Sustainable Practices:

Provide Actionable insights for improving crop rotation, resource allocation, and soil
conversation to ensure long-term agricultural sustainability.

5. Enable Decision-Making:
Equip farmers, agronomists, and policymakers with accurate and easy-to-interpret
data for better resource management and enhance agricultural productivity.
This Project aims to bridge the gap between traditional farming practices and modern
technology, fostering sustainable agriculture while improving the economic resilience of
farmers.

4|Page
2. Literature Review

2.1Challenges in Agricultural Monitoring and Land Assessment

Modern agriculture faces an array of challenges that impede efficient monitoring, land
assessment, and sustainable practices. Key issues include inconsistent access to reliable
monitoring tools, variability in soil and vegetation conditions, and difficulties in large-
scale analysis due to resource constraints.

Inaccessibility to Accurate Monitoring Tools

According to a report by the International Food Policy Research Institute (IFPRI),
smallholder farmers in developing regions often lack access to sophisticated tools that
enable real-time assessment of land health and productivity. Manual assessment methods
remain prevalent, which are prone to errors, inconsistencies, and delays in decision-
making. These constraints contribute to inefficiencies in resource usage and reduce crop
yields [2].
Spatial and Temporal Variability in Soil and Vegetation
A study conducted by the University of Illinois emphasizes the spatial heterogeneity of
soil characteristics across different regions, which affects nutrient availability, water
retention, and crop suitability. Temporal variability, influenced by seasonal and climatic
changes, further complicates monitoring efforts. Without real-time, location-specific
data, it is difficult for farmers to make informed decisions about fertilization, irrigation,
or land treatment [3].

Resource Constraints for Large-Scale Monitoring

The Food and Agriculture Organization (FAO) outlines how limited financial, technical,
and human resources restrict the adoption of modern agricultural monitoring systems in
many countries. Satellite data, drones, and IoT sensors offer promising solutions, but
high setup and operational costs create barriers to their widespread adoption. This
hinders the development of scalable and efficient land assessment frameworks,
particularly in rural areas [4].

5|Page
2.2 Advancements in Remote Sensing and Geospatial Analytics

Recent years have seen rapid advancements in remote sensing and geospatial data
analysis, enabling precise monitoring of agricultural conditions at scale. These
technologies enhance the ability to detect land degradation, monitor vegetation indices,
and inform climate-resilient agricultural practices.

Use of Satellite Imagery for Vegetation and Soil Monitoring

Research from the European Space Agency (ESA) demonstrates that high-resolution
satellite imagery can effectively monitor vegetation health and soil characteristics over
time. Sentinel-2, in particular, provides multispectral data suitable for calculating indices
such as NDVI (Normalized Difference Vegetation Index), NDMI (Normalized Difference
Moisture Index), and GCI (Green Chlorophyll Index). These indices are used to detect
drought, nutrient deficiencies, and crop stress [5].

Application of Google Earth Engine (GEE)

Google Earth Engine has emerged as a leading cloud-based platform for large-scale
geospatial analysis. According to Gorelick et al. (2017), GEE facilitates real-time
processing of petabytes of satellite imagery and geospatial datasets, empowering
researchers to analyze land cover changes, vegetation health, and environmental trends
without the need for local high-performance computing infrastructure [6].

Integration of AI with Remote Sensing

Studies from Stanford University show how machine learning models, when integrated
with satellite and aerial imagery, can predict crop yields and classify land usage with high
accuracy. These models learn from historical patterns and current environmental inputs,
supporting precision agriculture and timely interventions [7].

6|Page
2.3 Review of Existing Geospatial Platforms and Index Calculators

A range of platforms and tools currently exist to support remote agricultural monitoring
through geospatial and index-based analytics. However, limitations in usability,
flexibility, and user access remain.

Crop Monitoring Platforms (e.g., CropIn, EOS Crop Monitoring)

Platforms like CropIn and EOS offer visualization tools for crop monitoring, using
satellite data and vegetation indices. These services often include NDVI, EVI (Enhanced
Vegetation Index), and LAI (Leaf Area Index) as key metrics. However, research by the
Indian Institute of Science (IISc) notes that such platforms are often commercial, limiting
access for small-scale farmers due to subscription fees or restricted APIs [8].

Government-Driven GIS Portals (e.g., Bhuvan, USDA’s CropScape)

Bhuvan (India) and CropScape (USA) are government-funded GIS portals that provide
satellite-derived data on land use, vegetation, and soil health. While comprehensive, their
user interfaces are often not intuitive, and they primarily serve researchers and
policymakers rather than individual farmers or startups. Usability studies from the Indian
Space Research Organization (ISRO) suggest the need for simplified, mobile-friendly
interfaces [9].

Index Calculation Tools (e.g., QGIS Plugins, SNAP Toolbox)

Free and open-source software like QGIS and ESA's SNAP offer plugins for calculating
vegetation indices from satellite images. While powerful, they require a steep learning
curve and are better suited for experts. A usability survey from Wageningen University
found that farmers prefer automated tools that abstract the complexity of remote sensing,
offering direct insights and recommendations [10].

7|Page
3. System Overview and Design
3.1System Architecture

The architecture of SoilSense is designed with a modular, scalable approach that

integrates modern web development frameworks, cloud-based geospatial processing,
and robust data management. The system comprises five layers: the frontend
interface, backend processing, Google Earth Engine (GEE) integration, and a
persistent database layer. Each component interacts seamlessly to deliver a smooth
user experience and real-time geospatial analytics.

3.1.1 Overall Architecture of SoilSense

SoilSense is structured as a multi-tier application to ensure clear separation of
concerns, easy maintainability, and high scalability. The system is primarily divided
into:
 User Interface Layer ([Link]): Allows users to register, log in, draw areas of
interest on a map, and visualize results.
 Backend Processing Layer ([Link] + Python): Manages user authentication,
receives geospatial data, triggers Earth Engine computations, and communicates
with the database.
 Google Earth Engine Layer: Performs heavy geospatial computations like
vegetation index calculation using satellite data (Sentinel-2, Landsat-8).
 Database Layer (MongoDB): Stores user data, calculation history, coordinates,
and computed results.
The communication between these layers is handled via RESTful APIs, ensuring that
the system remains loosely coupled yet highly cohesive.

Fig 1: Overall architecture of Soil sense

8|Page
3.1.2 Frontend ([Link]) Interface Layer
The frontend of SoilSense is built using [Link], which provides a dynamic,
responsive, and user-friendly experience. Major features of this layer include:
 Interactive Map Drawing: Using libraries like react-leaflet, users can draw
polygons on a map to select their area of interest.
 User Authentication: Provides secure login and registration functionality
through JWT-based authentication.
 Dashboard and Result Display: Users can view previously computed results,
revisit past areas, and explore data visualizations using integrated chart libraries
(e.g., [Link], Recharts).
 Data Request Handling: Submits area coordinates and year range selections to
the backend through HTTP POST requests.
This layer is designed with a modern UI/UX approach, ensuring accessibility for
users with varying levels of technical expertise.

9|Page
3.1.3 Backend ([Link] + Python) Processing Layer
The backend is powered by [Link] for routing and Python for computation. This
hybrid structure allows for a flexible and performance-efficient pipeline.
 [Link]: Handles API endpoints for authentication (/auth), user data (/user), and
calculations (/calculate). It manages secure sessions and user roles.
 Python (calculate_indices.py): Invoked by [Link] to perform geospatial
calculations using Earth Engine. It calculates vegetation indices like NDVI,
NDMI, NDSI, GCI, EVI, AWEI, and LST.
 Error Handling & Validation: Validates inputs like polygon shapes, date
ranges, and handles server-side exceptions to ensure system stability.
This dual-language approach benefits from [Link]'s fast I/O and Python’s strong
geospatial and scientific processing capabilities [11].

3.1.4 Google Earth Engine (GEE) Integration Layer

Google Earth Engine (GEE) serves as the analytical engine for SoilSense. It provides
access to petabytes of satellite data and advanced geospatial analysis tools. Key tasks
performed in this layer include:
 Satellite Data Access: Fetches imagery from Sentinel-2 and Landsat-8
collections [12].
 Index Calculations: Computes multiple indices such as NDVI, NDMI, and LST
across selected timeframes[13].
 Temporal Aggregation: Aggregates values annually to provide trends over
selected year ranges.
 GeoJSON Area Filtering: Clips imagery based on user-defined GeoJSON areas,
ensuring precise region-specific calculations.
The integration is implemented using the Earth Engine Python API, which is
triggered server-side and sends the final results back to the backend for storage and
display.

3.1.5 Database Layer (MongoDB)

SoilSense uses MongoDB as its primary database solution due to its flexibility with
document-based data and ease of integrating with JavaScript-based stacks.
 User Data: Stores login credentials, profile information, and authorization
tokens.
 Calculation Records: Stores each calculation session, including area
coordinates, selected years, computed index values, and interpretation notes.
 Historical Retrieval: Allows users to revisit and reload previous calculations
with associated geospatial context.
 Scalability: The NoSQL structure makes it easy to extend data models as new
indices or features are added.
MongoDB’s schema-less nature and JSON-like documents make it particularly well-
suited to storing GeoJSON and user-driven input data [14].

10 | P a g e
11 | P a g e
3.2 Functional Workflow of SoilSense
The functional workflow of SoilSense outlines the step-by-step operations from the
moment a user interacts with the system to the final result visualization and storage. The
workflow is engineered to be intuitive, efficient, and informative, ensuring that even
users with limited technical knowledge can extract meaningful insights about soil health
and land conditions.

3.2.1 Area Selection and Index Computation via GEE

The first step in the SoilSense workflow involves the user selecting a geographical
area of interest using an interactive map embedded in the frontend interface.
 User Interaction: Users draw polygons directly on a Leaflet-based map interface
using tools from react-leaflet-draw [15].
 GeoJSON Export: Once drawn, the area is converted to GeoJSON format [16],
which represents the selected spatial boundary.
 Custom Input: Users also specify a year range for which they want to compute
the vegetation and land indices.
 Validation: The frontend validates the shape and size of the selected area to
ensure compatibility with the Earth Engine API limitations [11].
This feature enables precise targeting of regions, empowering users to analyze fields,
farmlands, or watersheds with high spatial accuracy.

Fig 1: Area selector for a new analysis

12 | P a g e
3.2.2 Data Retrieval and Index Computation via GEE
Once the area is selected, the frontend sends the GeoJSON coordinates and selected
years to the backend, where computation is handled by Google Earth Engine (GEE).
 Backend Triggering: The [Link] backend calls a Python script
(calculate_indices.py) which interfaces with GEE [14].
 Satellite Dataset Access: GEE fetches imagery from Sentinel-2 and Landsat-8
repositories based on the provided date range [12].
 Index Calculations: The system calculates multiple indices including:
o NDVI (Normalized Difference Vegetation Index)
o NDMI (Moisture Index)
o NDSI (Snow/Soil Index)
o GCI (Chlorophyll Indicator)
o EVI (Enhanced Vegetation Index)
o AWEI (Water Index)
 Temporal Averaging: The indices are averaged on a per-year basis to observe
long-term trends [17].
This layer transforms raw satellite data into structured, interpretable information
aligned with precision agriculture goals.

Fig 2: Region of Interest used to the plot in testing (Bhopal 50km radius)

Fig 3: Multispectral band data of the respective ROI (Region of Interest) here Bhopal
radius : 50km

13 | P a g e
3.2.3 Result Visualization and Interpretation
After the Earth Engine processing is complete, the backend receives computed
results and returns them to the frontend for real-time visualization.
 Graphical Representation: Each index is visualized using dynamic charts (e.g.,
bar graphs, line graphs) created using libraries like Recharts or [Link] [18].

Fig 4: Example of Graphical Representation (2022 – 24 NDVI Trend)

Fig 5: Example of Graphical Representation (2022 – 24 NDMI Trend)

14 | P a g e
Fig 6: Example of Graphical Representation (2022 – 24 NDSI Trend)

Fig 7: Example of Graphical Representation (2022 – 24 GCI Trend)

Fig 8: Example of Graphical Representation (2022 – 24 EVI Trend)

15 | P a g e
 Fig 9: Example of Graphical Representation (2022 – 24 AWEI Trend)

 Year-wise Comparison: The graphs clearly show changes across the selected
year range, helping users detect patterns in vegetation, moisture, or heat levels.
 Interpretative Text: Alongside graphs, the system provides automated
interpretative summaries explaining what the index values [19] suggest (e.g., "
NDVI 0.678 indicates dense healthy vegetation.").
 Mobile-Responsive Layout: All visualizations are designed to work seamlessly
on both desktop and mobile platforms [20].
This step translates complex geospatial data into user-friendly visual insights,
enhancing decision-making for farmers, researchers, and policymakers.

16 | P a g e
3.2.4 User Authentication and Calculation History
To enhance personalization and track historical insights, SoilSense includes user
authentication and stores a detailed calculation history.
 JWT-Based Login/Registration: Users create accounts and log in securely
using JSON Web Tokens (JWT) [21].
 User-Specific Data: Each calculation is associated with the authenticated user,
ensuring that data remains private and organized.
 Dashboard Access: Users can access a dashboard displaying past calculations,
timestamps, map previews, and index summaries.
 Revisit Functionality: A "revisit" feature allows users to re-open past analyses
and view previous results without triggering a re-calculation [22].
This ensures continuity in research and allows users to compare results over time or
across different locations.

Fig 10: Representing User dashboard after login

17 | P a g e
3.2.5 Data Persistence and Retrieval from Database
All input, output, and user data is stored securely in a MongoDB database to support
retrieval, performance tracking, and long-term use.
 GeoJSON Storage: User-drawn area boundaries are saved in JSON format for
future reuse.
 Computed Results: Index values, year ranges, and interpretation summaries are
stored alongside the original user request.
 Efficient Retrieval: The backend provides endpoints to query previous
calculations, enabling fast retrieval of data without needing to reprocess it.
 Scalability: MongoDB’s schema-less design allows new indices or metadata to
be added without disrupting existing functionality.
This layer ensures that SoilSense serves not only as a real-time tool but also as a data
archive that supports comparative analysis over multiple sessions and seasons.

18 | P a g e
3.3 Features and Outputs

SoilSense offers a robust set of features designed to streamline the process of land
monitoring, soil health analysis, and geospatial index computation. The platform
leverages modern web technologies and advanced Earth observation data to deliver
meaningful outputs that aid in sustainable agricultural planning and land resource
management.

3.3.1 Interactive User Interface

One of the defining aspects of SoilSense is its intuitive and interactive frontend
interface, built using [Link]. This interface plays a crucial role in ensuring that
users from both technical and non-technical backgrounds can navigate and utilize the
platform effectively.
 Map Drawing Tools: Users can easily draw polygons on a map to select areas of
interest using drag-and-drop tools. The interface is powered by Leaflet and react-
leaflet-draw, providing smooth interaction and accurate geolocation capabilities.
 Responsive Design: The application is fully responsive, meaning it adapts
seamlessly to desktop, tablet, and mobile screens, making it accessible to users in
the field.
 Dropdowns and Year Selection: Users can choose specific years or a range of
years for which they want to analyze environmental indices, enabling temporal
comparisons.
 Real-Time Feedback: The UI displays loading indicators, success messages, and
error prompts to improve user experience during each step of the analysis
workflow.
 Dashboard and Navigation: A user-friendly dashboard organizes historical
calculations, with clickable tiles that lead to full reports and visualizations from
past sessions.

This interactive interface makes the complex process of geospatial analysis

approachable and visually engaging.

19 | P a g e
3.3.2 Vegetation and Other Index Trends Over Time
The primary outputs of SoilSense are visualizations and interpretations of various
environmental indices derived from satellite imagery. These trends help users
understand long-term changes in vegetation health, water content, and land surface
temperature across the selected area.

 NDVI (Normalized Difference Vegetation Index): Reflects overall vegetation

health and biomass density. Trends can reveal crop cycles, deforestation, or land
degradation [23].
 NDMI (Normalized Difference Moisture Index): Indicates water content in
vegetation, helping identify drought stress or irrigation issues [24].
 NDSI (Normalized Difference Snow Index): Useful in areas with seasonal
snow or distinguishing bare soil vs. vegetation [25].
 GCI (Green Chlorophyll Index): Estimates chlorophyll content, providing
insight into plant nutrition and photosynthetic activity [26].
 EVI (Enhanced Vegetation Index): Offers more precise vegetation readings in
areas with dense canopy or atmospheric distortion [27].
 AWEI (Automated Water Extraction Index): Highlights surface water bodies
and helps detect flooding or drainage issues [28].
Each index is visualized with:
 Graphical Charts: Year-wise bar or line graphs generated using Recharts [18].
 Interpretive Texts: Summary explanations that describe what the trends suggest
in practical farming or environmental terms [19].
 Downloadable Reports: Users can export the results for offline use, sharing
with agronomists, or research documentation.

Together, these outputs offer a comprehensive overview of land conditions across

time, enabling data-driven decisions for farming, conservation, or environmental
assessment.

20 | P a g e
3.4 Individual Contributions

Shiv, Khushal
Main Role: Presentation (PPT)
Sub Role: Frontend

 Created the Project Presentation

o Designed and built the final presentation summarizing the application goals,
technologies used, functionality, and architecture.
o Explained the backend–frontend integration and usage of Google Earth Engine
(GEE).

 Frontend Contributions:
o Assisted with component styling (*.css files like [Link], [Link],
[Link], etc.).

 Helped design the login/signup pages:

o [Link], [Link]

 Worked on layout and structure of:

o [Link] – Displays user’s previous calculations.
o [Link] – Navigation across the app.
o [Link] – View-only mode for previous results.

Pratham, Nilesh
Main Role: Research Paper
Sub Role: Backend

 Research-Based Contributions:
o Researched vegetation and water indices, including NDVI, EVI, NDMI, NDSI,
etc.
o Compiled a paper explaining the scientific relevance of each index and its
interpretation in soil/water analysis.

 Backend Contributions:
o Helped build and organize backend routes in gee_backend/routes:
 [Link]: User authentication.
 [Link]: Route to trigger satellite index calculation.

 Contributed to creating:
o models/[Link] – Schema for storing each user’s calculations.
 models/[Link] – Schema for storing user credentials.
 middleware/[Link] – JWT authentication middleware.

21 | P a g e
Anshul, Viral
Main Role: Frontend
Sub Role: Database

 Frontend Contributions:
o Developed map interaction interface in:
 [Link] – Leaflet-based drawing tool for selecting regions.
 [Link] – Small map component showing preview for previous
calculations.

 Implemented graph rendering and visualization:

o [Link] – Includes graph components for yearly index display.
o Used [Link] (or similar) to visualize trends over time.

 Database Integration:
o Worked with storing results in MongoDB from the frontend using:
o [Link] – Contains API calls to backend (e.g., login, register, calculate).
o Contributed to managing data flow between the frontend and the backend after
calculation results are returned.

Devdeep
Main Role: Backend
Sub Role: Database

 Core Backend Developer:

o Built the server using:
 gee_backend/[Link] – Entry point for Express backend.
o Integrated routes:
 routes/[Link], routes/[Link]
o Created Python–Node bridge:
 Connected calculate_indices.py to Node using child_process.spawn().
 Executed Python script with user-input GeoJSON and year range.

 Google Earth Engine & Python:

o Implemented calculate_indices.py:
 Fetches Sentinel and Landsat satellite images.
 Calculates indices like NDVI, NDMI, NDSI, GCI, AWEI, EVI over years.

 Database Work:
o Used Mongoose schemas ([Link], [Link]) to store:
 User credentials and JWT.
 Calculation results along with the selected region and time span.

22 | P a g e
 Conclusion
The integration of modern technologies into agriculture has become a necessity in addressing
the growing challenges of food security, resource scarcity, and climate change. The satellite-
based NDVI detection systems demonstrate how innovative solutions can transform
traditional farming practices into data-driven, sustainable processes.
The use of satellite detection for agricultural land fertility using NDVI illustrates the power of
remote sensing in large-scale farming. By leveraging multispectral satellite imagery, NDVI
enables precise monitoring of vegetation health, soil fertility, and land classification. This
scalable technology ensures efficient resource management, early detection of crop stress,
and optimized land use planning, benefiting both smallholder and commercial farmers.
Technology can create measurable impacts on productivity, sustainability, and profitability in
agriculture. With continued advancements and widespread adoption, these systems have the
potential to revolutionize farming practices worldwide, ensuring food security and
environmental stewardship for future generations.

23 | P a g e
 References
1. Government of India. (2023). Agricultural Statistics at a Glance 2023. Ministry of
Agriculture & Farmers Welfare. 1-250. [Link]
2. International Food Policy Research Institute. (2020). Smallholder Access to Agricultural
Technology. IFPRI, Washington, D.C.
3. Jayne, T.S., Chamberlin, J., & Headey, D.D. (2019). Spatial heterogeneity of soil
properties and its implications for land productivity. Nature Sustainability, 2(10), 805–
813.
4. Food and Agriculture Organization of the United Nations. (2019). Digital Technologies in
Agriculture and Rural Areas – Status Report. FAO, Rome.
5. European Space Agency. (2021). Sentinel-2 for Agriculture: A Technical Guide. ESA.
6. Gorelick, N., Hancher, M., Dixon, M., Ilyushchenko, S., Thau, D., & Moore, R. (2017).
Google Earth Engine: Planetary-scale geospatial analysis for everyone. Proceedings of
the National Academy of Sciences, 114(49), 13108–13113.
7. Karpathy, A., & Fei-Fei, L. (2016). Deep Learning for Agriculture: Crop Yield Prediction
from Remote Sensing. Stanford University Technical Report.
8. Indian Institute of Science. (2020). Usability Study on Commercial Crop Monitoring
Platforms. IISc Research Brief.
9. Indian Space Research Organisation. (2021). User Experience Report: Bhuvan Geoportal.
ISRO Publication.
10. Wageningen University. (2020). Usability and Adoption of Open-Source Remote Sensing
Tools in Precision Agriculture. WUR Reports.
11. Gorelick, Noel & Hancher, Matt & Dixon, Mike & Ilyushchenko, Sergey & Thau, David
& Moore, Rebecca. (2017). Google Earth Engine: Planetary-scale geospatial analysis for
everyone. Proceedings of the National Academy of Sciences, 114(49), 13108–13113.
10.1073/pnas.1707601114.
12. Drusch, M., Del Bello, U., Carlier, S., Colin, O., Fernandez, V., Gascon, F., ... &
Bargellini, P. (2012). Sentinel-2: ESA's Optical High-Resolution Mission for GMES
Operational Services. Remote Sensing of Environment, 120, 25–36.
10.1016/[Link].2011.11.026.
13. Rouse, J.W., Haas, R.H., Schell, J.A., & Deering, D.W. (1974). Monitoring vegetation
systems in the Great Plains with ERTS. NASA SP, 351, 309.
14. Chodorow, Kristina. (2013). MongoDB: The Definitive Guide: Powerful and Scalable
Data Storage. 2nd Edition. O’Reilly Media. ISBN: 9781449344689.
15. Folium, React-Leaflet-Draw, and Leaflet. (2023). Leaflet draw tools and geospatial
rendering. Retrieved from [Link]
16. Butler, H. et al. (2008). The GeoJSON format: a community-driven open geospatial
standard. Open Geospatial Consortium. [Link]
17. Panday, P. K., & Ghimire, B. (2019). Examining NDVI and land surface temperature
trend in Kathmandu Valley (Nepal). Remote Sensing Applications: Society and
Environment, 14, 105–113. 10.1016/[Link].2019.02.003
18. Recharts Team. (2024). Recharts: A composable charting library built on React
components. Retrieved from [Link]

24 | P a g e
19. Jha, C. S., Goparaju, L., & Raghubanshi, A. S. (2020). Vegetation indices for assessing
health and productivity of land. Ecological Indicators, 117, 106546.
10.1016/[Link].2020.106546
20. Marcotte, Ethan. (2011). Responsive Web Design. A List Apart, 306.
[Link]
21. Jones, M., Bradley, J., & Sakimura, N. (2015). JSON Web Token (JWT). IETF RFC 7519.
[Link]
22. Kothencz, Gyula & Kiss, Márton & Petutschnig, Andreas. (2017). Hot spots for
improvements: Where to implement new green spaces? 1–4. 10.1109/UIC-
ATC.2017.8397576
23. Pettorelli, N. (2013). The Normalized Difference Vegetation Index. Oxford University
Press.
24. Gao, B.-C. (1996). NDWI—A normalized difference water index for remote sensing of
vegetation liquid water from space. Remote Sensing of Environment, 58(3), 257–266.
[Link]
25. Hall, D. K., Riggs, G. A., & Salomonson, V. V. (1995). Development of methods for
mapping global snow cover using moderate resolution imaging spectroradiometer data.
Remote Sensing of Environment, 54(2), 127–140.
26. Gitelson, A. A., & Merzlyak, M. N. (1998). Remote sensing of chlorophyll concentration
in higher plant leaves. Advances in Space Research, 22(5), 689–692.
27. Huete, A. R., et al. (2002). Overview of the radiometric and biophysical performance of
the MODIS vegetation indices. Remote Sensing of Environment, 83(1–2), 195–213.
28. Feyisa, G. L., Meilby, H., Fensholt, R., & Proud, S. R. (2014). Automated Water
Extraction Index: A new technique for surface water mapping using Landsat imagery.
Remote Sensing of Environment, 140, 23–35. [Link]
29.

25 | P a g e
 BIO-DATA

Name: Anshul Ajit Dalvi

Reg. No: 22BCE10799

Email: anshuldalvi2022@[Link]

DOB: 30/03/2004

Field of Study: Computer science

Key skill: Data analysis

Degree: [Link]

Name: Dev Deep Narayan

Reg. No: 22BEC11141
Email: devdeepnarayan@2022@[Link]
DOB: 29/08/2004
Field of Study: Computer Science
Key skill: Data Analysis
Degree: [Link]

Name: Khushal Jain

Reg. No: 22BCE10292

Email: khushaljain2022@[Link]
DOB: 29/11/2003
Field of Study: Computer Science
Key Skill: Web Developer

Degree: [Link]

26 | P a g e
 Name: Nilesh Pant

Reg. No: 22BEC10014

Email: nileshpant2022@[Link]
DOB: 24/08/2004
Field of Study: Electronics and communication Eng.
Key Skill: Core electronics
Degree: [Link]

Name: Pratham Amit Singh

Reg. No: 22BCE10959

Email: prathamamitsingh2022@[Link]
DOB: 23/05/2005

Field of Study: Computer Science

Key Skill: Data Structures
Degree: [Link]

Name: Shiv Prakash Mandhani

Reg. No: 22BAI10172

Email: shivprakashmandhani2022@[Link]
DOB: 17/12/2003
Field of Study: Computer Science(AIML)
Key Skill: Data Analysis

Degree: [Link]

27 | P a g e
 Name: Viral Sancheti
Reg. No: 22BCE11406
Email: Viralsancheti2022@[Link]
DOB: 11/05/2005
Field of Study: Computer science
Key Skill: Artificial intelligence
Degree: [Link]

28 | P a g e
SATELLITE DETECTION FOR AGRICULTURAL LAND AND FERTILITY USING
NDVI – 4/14/2025
Phae II Report
Submitted by:
22BCE10799 Anshul Ajit Dalvi

22BCE11141 Dev Deep Narayan

22BCE10292 Khushal Jain

22BEC10014 Nilesh Pant

22BCE10959 Pratham Amit Singh

22BAI10172 Shiv Prakash Mandhani

22BCE11406 Viral Sancheti

29 | P a g e