PLANT DISEASE DETECTION

DESIGN THINKING MODULE 002 PROJECT REPORT

Group-B28

Submitted by

TRINATH REDDY M(2021BCSE07AED263)

SAI YESWANTH REDDY D(2021BCSE07AED348)

MANOJ B N(2021BCSE07AED079)

In partial fulfilment for the award of the degree of

BACHELOR OF TECHNOLOGY

Under the Supervision of

Dr. ABRAHAM GEORGE

Dr.ATUL

Dr. CHITRA KIRAN

ALLIANCE COLLEGE OF ENGINEERING AND DESIGN

BENGALURU

NOVEMBER – 2023
 DECLARATION

This is to declare that the report titled “PLANT DISEASE DETECTION” has been made for the partial
fulfilment of the Course Bachelor of Technology in Computer Science and Engineering, under the
guidance of Dr. ABRAHAM GEORGE, Dr.ATUL, Dr. CHITRA KIRAN. I confirm that this report
truly represents my work undertaken as a part of my DTM 002 project work. This work is not a replication
of work done previously by any other person. I also confirm that the contents of the report and the views
contained therein have been discussed and deliberated with the faculty guides.

NAME REG.NO SIGNATURE

TRINATH 2021BCSE07AED550

YESWANTH 2021BCSE07AED348

MANOJ 2021BCSE07AED079

i
 CERTIFICATE
This is to certify that the project work entitled “PLANT DISEASE DETECTION” is the bonafide work
done by Trinath reddy (2021BCSE07AED550) submitted in the partial fulfilment of the requirements for
the award of the degree Bachelor of Technology in Computer science Engineering, Yeswanth reddy
(2021LMEC07AED348) submitted in the partial fulfilment of the requirements for the award of the
degree Bachelor of Technology in Mechanical Engineering. Manoj K(2021BCSE07AED079)submitted in
the partial fulfilment of the requirements for the award of the degree Bachelor of Technology in Computer
science Engineering, during the year 2023.

Dr.Abraham George Dr.Atul Dr. Chitra Kiran

Supervisor Supervisor Supervisor

Dept. of CSE Dept. of ME Dept. of ME

External Examiners:

1. Name: Signature

2. Name: Signature

3. Name: Signature

ii
 ACKNOWLEDGEMENT

The satisfaction that accompanies the successful completion of the task would be put incomplete without
the mention of the people that made it possible, whose constant guidance and encouragement crown all the
efforts successfully. I am very thankful to my Supervisors , Dr. ABRAHAM GEORGE, Dr.ATUL, Dr.
CHITRA KIRAN for their sustained inspiring guidance and cooperation throughout the process of this
project. Their wise counsel and valuable suggestions are invaluable. I would like to thank Dr. Abraham
George, Head of the Department and Dr. Reeba Korah, Dean for her encouragement and cooperation at
various levels of Project. I would like to extend my special thanks to Dr. Saswati Debnath, Associate
Professor and DTM Co-ordinator of Alliance College of Engineering and Design. I avail this chance to
express my deep sense of gratitude and hearty thanks to the Management of Alliance University, for
providing world class infrastructure, congenial atmosphere, and encouragement. I express my deep sense
of gratitude and because of the teaching and non-teaching staff at our department who stood with me
during the project and helped me to make it a successful venture. I place highest regards to my parents, my
friends and well-wishers who helped plenty in making the report of this project.

TRINATH REDDY M(2021BCSE07AED550)

SAI YESWANTH REDDY D(2021BCSE07AED348)

MANOJ B N(2021BCSE07AED079)

iii
 LIST OF FIGURES

SL.NO FIGURE NAME CAPTION PAGE

NO
1 Plant Leaf Sample of Plants 01

2 Plant Leaf Tomato and leaf of tomato 04

3 CNN Model Architecture 07

4 Model overview Flow chart of proposed model 09

iv
 ABSTRACT

The process of identifying medicinal plants is complicated and multidimensional, and it is crucial to both
current scientific study and conventional medical practices. There are many obstacles in the way of this
search, and misidentification is one of the biggest ones. Misidentification can result in mistakes in
traditional use, scientific research, and medicinal applications due to variables including physical
similarity, regional variations, and poor training. The ramifications of misidentification are examined in
this abstract, including the potential loss of traditional knowledge and the ineffectiveness of therapy.

Education and training, standardization of identification techniques, cooperation between researchers and
traditional practitioners, and the moral integration of technology are all vitally needed to lessen this
problem. In the end, addressing misidentification promotes the sustainable and responsible use of these
priceless natural resources by ensuring the correctness of plant categorization, supporting the preservation
of traditional knowledge, and enhancing the dependability of medicinal plant research.

Accurate identification of medicinal plants is crucial in Ayurvedic medicine. Experts on humans identify
plants based on their visual characteristics and fragrance. Bad things can happen if medicinal plants are
incorrectly identified. Visual morphological traits including leaf and blossom form, color, and texture may
be used to automatically identify plants. The form, color, and texture of a leaf are the primary
characteristics needed to identify a medicinal plant. Deterministic characteristics for species identification
are present in the color and texture of the leaf on both sides.

Therefore, a vision-based approach may help in the quick and precise identification of medicinal plants by
both researchers and the public. To detect herb plants, this study suggests using intelligent vision and an
autonomous convolutional neural network (CNN). A CNN block for feature extraction and a classifier
block for categorizing the retrieved features make up the suggested Deep Learning (DL) model.

A comprehensive strategy is required to handle this issue, one that incorporates modern technologies like
machine learning and deep learning, standardizes identification processes, and launches education and
training programs. Improving the precision of medicinal plant identification can fortify the connection
between conventional knowledge and contemporary research, guaranteeing the sustained utilization of
these priceless botanical assets for the advancement of world health and welfare.

v
 CONTENTS

Page no.
Declaration …………………… i
Certificate …………………… ii
Acknowledgement …………………… iii
List of Figures ………………….... iv
Abstract …………………… v

Chapter 1: INTRODUCTION 01 - 03
1.1 Introduction
1.2 Context and Problem Statement
1.3 Technological Foundations
1.4 Significance of the Project

Chapter 2: IDENTIFYING THE PROBLEM 04 - 05

2.1 Introduction
2.2 Misidentification
2.3 Depends on Botanical Expertise
2.4 Public Awareness
2.5 Traditional Knowledge

Chapter 3: OBJECTIVES OF THE THESIS 06 - 09

3.1 Project Aim
3.2 Constructing a Deep Learning Model
3.3 Gathering and Preparing Data
3.4 Improvement of Model Performance
3.5 Combining Mobile Apps with Integration

vi
Chapter 4: PROCESS/METHODOLOGY 10- 16
4.1 Identify the needs of customer and users through field visits/surveys.
4.2 Translate the needs of customer into product.
4.3 Specification of the product
4.4 Designing the solution for required customer needs

Chapter 5: Literature Survey 17

Chapter 6: CONTRIBUTION TO RESEARCH 18
5.1 Individual contribution

Chapter 7: CONCLUSION AND FUTURE WORK 19

REFERENCES 20

vii
Chapter 1
INTRODUCTION

________________________________________________________

1.1 Introduction

In the field of agriculture, plant health and vitality are critical to guaranteeing the security of the world's
food supply. Plant diseases, on the other hand, represent a serious risk to crop yields and could result in
severe financial losses as well as food shortages. The implementation of successful mitigation methods
necessitates the timely detection and correct diagnosis of these diseases. Rapid technological
breakthroughs, especially in the areas of deep learning, machine learning, and image processing, present a
chance to completely transform the way we manage and monitor plant health.

The construction of an advanced plant disease detection system using Python, a programming language
that is popular and versatile, is the main emphasis of this work. Using Python frameworks and libraries to
their full potential, our goal is to develop a complete solution that combines cutting-edge deep learning
models, conventional machine learning algorithms, and image processing approaches. The main objective
is to provide agronomists and farmers with a tool that can quickly and precisely identify plant illnesses,
enabling prompt disease management and intervention.

This collection probably includes a wide variety of plant species, each of which has unique characteristics
that are essential for precise identification. The extensive information included in the dataset may be
accessed by using machine learning algorithms and picture recognition techniques.

1
1.2 Context and Problem Statement

The identification of medicinal plants is important in several fields, including pharmacology, conservation
biology, and traditional medicine. Medicinal plants have been utilized for thousands of years to treat a
wide range of ailments and have greatly aided the pharmaceutical industry. Traditional methods of plant
identification by scientists rely on aspects such as flower shape, leaf morphology, and other botanical
characteristics, and are carried out through manual observation. However, these methods are laborious,
arbitrary, and often require specialized knowledge.

2
With the growing global interest in herbal medicine and the sustainable use of natural resources, accurate
and efficient methods of identifying medicinal plants are becoming more and more important. The loss of
traditional knowledge about plant identification also makes it more difficult to preserve and use the
therapeutic properties of various plant species.

At issue are the inefficiency and subjectivity of traditional methods for identifying medicinal plants.
Manual plant identification is labor-intensive, technical, and prone to error due to individual perception
discrepancies. With the increasing demand for herbal treatments and the growing threat to biodiversity, a
systematic and objective approach to identify medicinal plants is desperately needed.

1.2 Technological Foundations

Advances in deep learning (DL), machine learning (ML), and computer vision constitute the primary
technical pillars supporting the identification of medicinal plants. Computer vision is a subfield of artificial
intelligence (AI) that plays a key role in automating the analysis of visual data and facilitating the feature
extraction of photos of plant specimens. Pattern recognition, classification, and feature extraction are all
done with machine learning techniques. Support vector machines (SVM), k-Nearest Neighbors, and
Random Forests are a few examples of these methods.

But the identification of medicinal plants has undergone a revolutionary breakthrough thanks to the
application of deep learning, specifically Convolutional Neural Networks (CNNs). CNNs are highly
effective at picture recognition tasks because they automatically extract hierarchical features from the
input, eliminating the need for human feature engineering.

1.4 Significance of the Project

The project's complex and cross-domain relevance lies in its use of machine learning and deep learning
techniques to the identification of medicinal plants. First off, the use of cutting-edge technology fills a
vital requirement for quick and precise plant identification, which has consequences for pharmacology,
traditional medicine, and biodiversity preservation. The initiative helps to preserve traditional knowledge
about medicinal plants by automating the identification process. This ensures that important information
about the qualities and applications of these plants is not lost in the face of dwindling expertise.

3
Chapter 2
IDENTIFYING THE PROBLEM
________________________________________________________
2.1 Introduction
At the interface of traditional medicine, pharmacology, and conservation biology, the identification of
medicinal plants is an essential task. Throughout history, people have turned to plants as primary sources
of healing for a wide range of diseases. This has greatly aided in the creation of medicines. However, there
are several obstacles to overcome to accurately identify medicinal plants, including ecological,
sociocultural, and taxonomic complexity. In order throw light on the difficulties encountered in the process
of identifying medicinal plants and to emphasize the need for creative solutions, this study attempts to
thoroughly investigate and evaluate these obstacles.

2.2 Misidentification

Misidentification of medicinal plants can occur for a number of reasons. Scientific complexity, species-to-
species physical similarity, and environmental variation all contribute to confusion. Dependence on
traditional knowledge, which differs among cultures, and a lack of standardized identification method raise
the possibility of misclassification even more.

The consequences of misidentification can be extensive. Relying on traditional medicine's misidentified

medicinal plants could lead to subpar results or, in the worst cases, ineffective treatments. In
pharmacological research, misidentification can skew results and hinder the development of accurate and
potent drugs. Misidentification could result in an unsustainable plant harvest, jeopardizing biodiversity and
ecological balance from a conservation perspective.

Fig 2. Tomato and Tomato leaf

4
The invisible leaves on both plants have a similar appearance. Therefore, we have to finish the image
recognition exercises.

2.3 Depends on Botanical Expertise

The diminishing corpus of botanical knowledge poses a significant problem in the field of medicinal
plants, where traditional knowledge is crucial to identifying, utilizing, and protecting the therapeutic
qualities of various botanical species. There is a serious fear that as interest in botany wanes, the centuries-
old knowledge held by knowledgeable herbalists and indigenous groups may be lost. This could make it
more difficult to find new uses for plants in medicine and put the sustainable management of plant
resources at risk.

2.4 Public Awareness

Increasing knowledge about medicinal plants includes not only their therapeutic uses but also the wider
ecological effects of sustainable practices. Increasing community responsibility increases the possibility of
supporting conservation efforts and guaranteeing the long-term survival of medicinal plant species.
Maintaining biodiversity and protecting the natural ecosystems that sustain these priceless resources
requires an understanding of the delicate balance that must be struck between harvesting and regeneration.

Public awareness campaigns that highlight the many uses and cultural significance of medicinal plants can
give people the information and confidence to make informed choices. Educational programs can draw
attention to the fact that the sustainable use of medicinal plants is not only a matter of individual health but
also a collective duty towards environmental stewardship.

2.5 Traditional Knowledge

Maintaining conventional knowledge about the characteristics and applications of plants is a crucial duty
that affects sustainable resource management and customs. There is a good chance that knowledge that has
been passed down through the ages will be lost if younger generations lose interest in picking up
antiquated skills. This decline poses a threat not only to the continuation of customs but also to the
precarious balance between ecological resources and human populations that has prevailed for many
generations.

5
Chapter 3
OBJECTIVES OF THE THESIS

________________________________________________________
3.1 Project Aim
A project named "PLANT DISEASE DETECTION" might aim to develop an automated system for to
detect plant disease identification that is reliable and efficient through the use of deep learning. The
principal aim is to leverage the capabilities of advanced computer methods to enhance the accuracy,
effectiveness, and comprehensibility of plant identification in the context of medicinal flora. By bridging
the gap between traditional methods of plant identification and modern technologies, the initiative hopes to
advance the fields of botany, pharmacology, and biodiversity protection. The objective is to create a
practical tool that will assist researchers, botanists, and individuals in accurately identifying medicinal
plants for a variety of applications, including environmental and therapeutic projects.

Additionally, it plans to look into the potential for transfer learning, which would allow the model to
benefit from past knowledge gained in other domains. This approach can improve the efficacy of the
identification procedure, especially when dealing with a large range of species of medicinal plants. In
addition, the project intends to offer an easy-to-use user interface that will facilitate deep learning system
usage for practitioners, scholars, and enthusiasts with varying levels of technical expertise.

3.2 Constructing a Deep Learning Model

The process of developing a deep learning model for medicinal plant identification is laborious. Gathering
a heterogeneous dataset is the first step in the process. To ensure consistency and enhance model
generalization, the dataset is then preprocessed using techniques like data augmentation. One popular
model architecture for extracting complex visual information from plant photos is the Convolutional
Neural Network (CNN).

The dataset is split into training, validation, and testing segments before the model is trained using an
appropriate loss function and optimizer. Hyperparameters are tweaked for best performance, and
regularization techniques are applied to lessen overfitting. The model is evaluated using metrics such as
accuracy and F1 score, in addition to a thorough analysis of the confusion matrix and error patterns. The
decision-making process of the model could be better understood by utilizing interpretability tools like

6
Grad-CAM. To guarantee the model's proper and moral integration into the identification of plant disease
applications, Documentation, continuous development, ethical issues, and deployment are crucial.
Collaboration and peer review also contribute to the advancement of the field.

Figure-3 CNN: Model Architecture

3.3 Gathering and preparing data

Care must be taken when gathering and arranging information about medicinal plants. The first step in the
process is to locate a large number of superb images that accurately represent various medicinal plants.
These images should show a variety of species, developmental stages, and environmental conditions to
ensure the model's robustness. Carefully choosing the collection means taking out any extraneous or
unnecessary photos and labeling each one with relevant information such as the kind of plant, what it
contains, and any potential medicinal uses.

An extensive dataset for medicinal plants must be created by carefully collecting and preparing data. The
first step in the process is to find a wide range of excellent photos that accurately depict various medicinal
plants. These images should feature a variety of species, developmental stages, and environmental

7
conditions to ensure the model's robustness. A carefully chosen dataset removes extraneous or pointless
images and annotates each one with relevant information such as the type of plant, the ingredients used,
and any potential medicinal advantages.

3.4 Improvement of Model Performance

To optimize the deep learning model's performance for medicinal plant identification, its parameters must
be adjusted to achieve high accuracy, precision, and recall. The model needs to be trained with the
previously chosen dataset, and hyperparameters like learning rates, batch sizes, and network topologies
need to be changed. Special attention should be given to addressing variations in shape, color, and texture
of medicinal plants, as these variations can have a significant impact on model performance.
Rotation, scaling, and flipping are examples of data augmentation techniques that can be used to improve
model generalization across a wide range of plant features. Regularization strategies like dropout can
prevent overfitting and boost the model's resilience. It is essential that the model be continuously evaluated
and validated on various datasets to ensure its effectiveness and suitability. Iterative model refining based
on validation findings is necessary to develop a deep learning model that functions well and can reliably
and accurately identify medicinal plants.

3.5 Combining Mobile Apps with Integration

When incorporating the deep learning model into a mobile application, a user-friendly interface has to be
made in order to increase accessibility and convenience for plant identification in the field. With the
application's aesthetically pleasing and intuitive interface, users should be able to snap and upload images
of therapeutic plants with ease. This interface is specifically designed for mobile devices. The deep
learning model should integrate smoothly into the application's backend, enabling real-time analysis of the
submitted photos. With proper training, the model can recognize plants.The recognized plant species,
pertinent details regarding their therapeutic qualities, and any necessary warnings or usage instructions can
all be seen on the user interface. Interactive elements like voice commands, augmented reality, and a plant
encyclopedia can be added to improve the user experience. In order to accommodate field scenarios with

8
limited connectivity, offline functionality can also be taken into consideration. Frequent model and
database updates can guarantee that the application remains up to date and keeps offering users who are
actively involved in medical plant identification on the go accurate and useful information.

3.6 Image Processing Techniques

The identification of medicinal plants using image processing techniques is an innovative approach at the
intersection of botany and technology. Scientists can analyze and interpret images of plant specimens to
extract key features like flower structures, leaf patterns, and overall shape by employing sophisticated
algorithms. As distinguishing markers for different plant species, these traits enable the development of
accurate and reliable identification models.
For this procedure to be automated and to allow for the rapid and scalable identification of medicinal
plants, deep learning architectures and other image recognition algorithms are necessary. This
technological advancement has many potential applications in pharmacology, ethnobotany, and traditional
medicine in addition to expediting the identification process. This methodology makes use of image
processing to allow for a more efficient.

9
Fig 4: Model overview: Flow chart of proposed system

10
Chapter 4
PROCESS/METHODOLOGY

________________________________________________________
4.1 Identify the needs of customer and users through field visits/surveys.

Surveys and field visits play a crucial role in identifying medicinal plants in diverse environments by
providing insight into the needs and challenges encountered by users. The results of a Google Forms
survey point to a significant problem: many respondents find it challenging to identify medicinal plants,
which can take up to 15 to 20 minutes on average, especially in places with a high biodiversity. This
significant time investment indicates a potential source of irritation and emphasizes the need for practical
and efficient solutions in the field of medicinal plant identification.

The poll also reveals an interesting trend: a sizable percentage of participants decide not to use
applications for medicinal plant identification, even when they are easily accessible. This finding raises the
question of whether the features offered by the available plant identification apps satisfy users'
requirements or preferences. To improve user adoption and satisfaction in the field of medicinal plant
identification, it is imperative to understand the factors that contribute to users' reluctance to use these
applications and tailor solutions that address their concerns.

The difficulties are greater in natural settings as respondents report more difficulty accurately identifying
medicinal plants. Urban congestion and limited access to medicinal plant species make these problems
worse, which emphasizes the need for innovative solutions that speed up the identification procedure.

The information gathered from surveys and fieldwork is invaluable for the creation of a medicinal plant
identification system. By putting the needs of the user first, this system may cut down on the time users
spend identifying plants, encourage the widespread use of plant identification apps, and offer customized
solutions to address the particular issues raised by different natural.

To save users time when identifying plants, more applications for plant identification should be used, and
solutions should be tailored to different scenarios. It is obvious that time-saving techniques, simple user
interfaces, and tailored strategies are required. The system is capable of effectively managing challenges
and accelerating the identification of medicinal plants by taking into account the preferences and actions of
its users.

11
4.2 Translate the needs of customer into product

Developing an innovative and user-focused mobile application is necessary to turn the expressed demands
of the market into a tangible product in the field of medicinal plant disease detection.The application will
significantly reduce the amount of time users spend searching up information on medicinal plants by
placing a high priority on usability. A well-designed user interface with visually appealing and intuitive
design elements is essential to enhancing the user experience overall.

The poll's results, which indicate that users are hesitant to use the current plant identification applications,
emphasize how important it is to match the functionality of the new application to users' actual needs and
preferences. This means closely analyzing the shortcomings in the existing applications and adding
features that directly address user concerns in order to increase adoption rates.

In urban settings where locating medicinal plants presents more challenges, the product ought to offer
tailored solutions. Using complex image identification algorithms that can manage variations in plant
appearance caused by developmental stages or environmental factors may be one way to achieve this.
Furthermore, given the constraints posed by a dearth of green spaces and urban congestion, the application
can offer location-based services to offer insights regarding local plant species.

The proposed mobile application for medicinal plant identification aims to transform the user experience
by responding to the requirements that have been discovered via surveys and on-site inspections. The
program prioritizes user comfort by featuring an easy-to-use interface that is visually appealing. This helps
users spend less time searching for information on a variety of medicinal plants. With the use of
sophisticated picture recognition algorithms, it can accurately identify plants in both urban and natural
environments by accounting for changes in form, color, and development phases.

The new application addresses the reported reluctance to use the current plant identification applications
by promoting alignment with user preferences. Beyond basic functionality, it offers a seamless and user-
friendly experience that encourages widespread adoption. The offline feature increases accessibility even
more by allowing users to utilize the app in locations with sporadic internet access, and regular updates
ensure the app's dependability and upkeep.

12
Customized solutions for urban issues are one of the key focus areas; the application provides location-
based services and details on native plant species. This user-centric smartphone application seeks to
become a comprehensive and indispensable resource for anyone seeking accurate and timely identification
of medicinal plants, in order to meet the diverse needs of users in a range of contexts.

The program recognizes the constraints imposed by a lack of open spaces and the hustle and bustle of the
city, going beyond simple identification. It provides intelligent information on native plant species by
combining location-based services. This feature optimizes the application's usefulness in densely
populated areas and enhances user experience by providing context-specific information tailored to the
urban environment.

In response to user feedback expressing concerns regarding the current plant identification applications,
the new application prioritizes a seamless and user-friendly experience. Beyond its fundamental features, it
offers an easy-to-use interface that encourages broader adoption. People who live in areas with
inconsistent internet access are taken care of and accessibility is further enhanced by the addition of an
offline option. Regular updates and maintenance guarantee the long-term dependability of the application
and enhancements, making it a solid resource for medicinal plant identification in a variety of urban and
natural settings.

13
4.3 Specification of the product

A sophisticated plant disease detection system is the novel idea being proposed, and it aims to
revolutionize the herbal medicine market. This product uses cutting-edge image processing techniques to
accurately identify and count the various kinds of medicinal plants found in a given area. The information
below provides more specifics about the primary features and functionalities of this cutting-edge interface.

 Interface and Experience for Users

The purpose of the plant disease detection is mobile application's user interface and experience is to ensure
accessibility, efficacy, and a seamless interaction by designing it with the needs of the user in mind. The
visually appealing layout of the intuitively designed interface prioritizes ease of use and simplified
navigation. Users can easily and enjoyably identify plants thanks to the application's user-friendly design.

Augmented reality (AR) capabilities enhance the user experience by enabling real-time information
overlays through the camera of the device. Users can visually identify and learn about nearby medicinal
plants thanks to this immersive capability, which not only boosts user engagement but also provides an
interesting and educational experience.

Personalization is an essential part of the user interface because it enables users to create distinctive
profiles with preferences, allowing them to tailor the experience to their own needs and interests. By
enabling hands-free operation, the voice command feature guarantees comfort during fieldwork or
exploration. In addition, the social media platform fosters a sense of community by encouraging users to
share their experiences, ask for assistance, and ask questions, all of which raise user engagement levels.

The primary objective of the design is to minimize the time users spend searching for plant information,
with regular updates to improve efficiency, accuracy, and speed. The app's offline features allow users to
use it uninterrupted in locations with spotty internet. The experience and user interface aim to provide a
thorough and user-friendly platform for effective medical plant identification, education, and community
engagement through these well considered design aspects.

14
  Distinctive characteristics

The identification features of the plant disease detection mobile application are designed to offer a precise
and fulfilling user experience. The sophisticated picture recognition algorithms at the heart of the
application allow for variations in color, shape, and stage of development to guarantee accurate
identification of medicinal plants. Users can confidently and accurately identify a wide range of plant
species in both urban and natural environments with the aid of this state-of-the-art technology.

The program goes beyond simple identification to provide a comprehensive encyclopaedia of plants. This
feature provides in-depth information on the medicinal properties, usage suggestions, and safety
precautions of every recognized plant. By addressing the challenges posed by different environments and
offering details about native plant species, location-based services enhance the identification process.

All things considered, the identification features are meant to give users a sophisticated tool that will help
them accurately identify medicinal plants while also increasing their understanding of each species they
have identified. This comprehensive approach ensures that users receive relevant and helpful information,
promoting a more informed and confident engagement with the field of medicinal plant identification.

 Convenient and Time Efficiency

The mobile application for plant disease detection is thoughtfully designed to maximize the user
experience, with an emphasis on speed and ease of use. Given how long it may take users to look up
information on medicinal plants, the program places a high priority on cutting down on the amount of time
they must spend doing so. Users can find accurate and up-to-date information quickly and easily thanks to
the application's well-designed user interface.

Regular updates enhance the application's responsiveness and speed, making it operate more quickly. The
app's offline feature is carefully integrated to offer uninterrupted access, taking into account the possibility
that users may be conducting research or fieldwork in areas with sporadic connectivity.

Ease is also prioritized in the application's user-centric design. The app enables users to effectively interact
and alter it according to their preferences by giving them the ability to customize their user profiles. The
ability to operate a device hands-free using voice command is particularly helpful for people who engage
in outdoor activities.

15
  Instructional Materials

The medical plant identification mobile application is a comprehensive educational tool meant to enhance
users' understanding and familiarity with medicinal plants, in addition to serving as a tool for species
identification. The app contains rich content, such as in-depth information on the recognized species of
medicinal plants. This content includes the identification of the plant along with details on its medicinal
properties, suggested use, and potential safety precautions.

In addition to factual information, the program offers instructional features like lessons and interactive
instructions. The careful integration of these resources empowers users and will gradually enhance their
plant identification skills. Through interactive learning opportunities, users of the app can expand their
knowledge of various medicinal plants, fostering more confident and informed interactions with their
environment.

The integration of the social platform further improves the educational component by fostering a sense of
community among users. This platform offers a forum for the exchange of opinions, experiences, and
requests for advice. Users can broaden their knowledge and gain understanding from the diverse
perspectives of a community of experts and enthusiasts through collaborative learning.

In summary, the application's educational resources for medicinal plant identification aim to provide users
with a thorough education that goes beyond simple identification. The app is an excellent resource for
anyone interested in learning more about therapeutic plants and their applications because it offers precise
information, tutorials, and a helpful community.

16
4.4 Designing the solution for required customer needs

In order to satisfy the stated consumer demands for the identification of plants disease, a comprehensive
solution must first be designed, which requires the development of a user-centric mobile application. The
program's user interface has been carefully created to offer easy navigation and simplicity of use. It is both
aesthetically pleasing and simple to use. By taking into account variations in growth stages, color, and
form, sophisticated image recognition algorithms are used to accurately identify medicinal plants,
increasing the reliability of the identification process.

One way to address the reported resistance to using the current plant disease identification applications is
to match features to user preferences. Augmented reality (AR) features use the device's camera to provide
users with an immersive, real-time experience that adds an interactive and educational component to the
identification process. A more comfortable and personalized experience is provided by voice command
capabilities and customized user profiles, which allow for hands-free use in a range of scenarios and
accommodate individual preferences.

By cutting down on the time users spend searching for plant information, the method prioritizes time
efficiency. Thanks to frequent updates and offline capabilities, users are ensured to have access to the most
relevant and up-to-date information, even in locations with inconsistent connectivity. The program's
comprehensive plant encyclopaedia gives users detailed information on the safety precautions,
recommended uses, and medicinal properties of known plant species.

Examples of instructional resources that improve user experience by helping people progressively get
better at identifying plants are tutorials and interactive guides. By facilitating the sharing of experiences,
viewpoints, and recommendations among users, the incorporation of a social media platform fosters a
sense of community and creates a collaborative learning environment.

In summary, the developed solution goes beyond simple identification by offering a comprehensive, user-
friendly, and educational platform for individuals wishing to identify and learn about medicinal plants.
Through its thoughtful design elements and functionalities, the solution seeks to empower users, promote
participation, and contribute to a more informed and confident contact with the world of medicinal plants.

17
Chapter 5
LITERATURE SURVEY

________________________________________________________
Prasanna Mohanty et al. proposed a method for plant disease detection through convolutional neural
network training in their paper “Deep learning for Image-Based Plant detection” [1]. The CNN model has
been trained to identify 14 distinct plant species as either healthy or diseased. The model’s accuracy with
the test set data was 99.35%. Although this is still preferable to a simple random selection model, accuracy
may be improved with a more varied training set. When applied to images sourced from credible websites,
the model attains an accuracy rate of 31.4%. Furthermore, alternative model or neural network training
variations may yield higher accuracy, paving the way for plant disease detection to become widely
accessible.

In the work “Leaf Disease Detection and Classification,” Malvika Ranjan et al. described a
technique for identifying plant diseases by using a photo of the afflicted leaf that was taken “using
Artificial Neural Network.” An Artificial Neural Network (ANN) needs to be trained by carefully
choosing feature values in order to discriminate between plant samples that are healthy and those
that are diseased. The ANN model has an 80% accuracy rate.

In their paper “Applying image processing technique to detect plant diseases,” Kulkarni et al.
detailed a methodology for early and accurate plant disease detection using artificial neural
networks (ANN) and a range of image processing techniques [4]. Because the suggested approach
is based on an ANN classifier for classification and a Gabor filter for feature extraction, it
produces better results, with a recognition rate of up to 91%.

A technique In the paper “Plant disease detection using CNN and GAN” [5], Emaneul Cortes
proposed a method for detecting plant diseases using Generative Adversarial networks.
Background segmentation is used to ensure precise feature extraction and output mapping. Gans

18
classification looks like a promising approach for classifying plant diseases, but background-
segmented data did not improve accuracy.

Chapter 6
CONTRIBUTION TO RESEARCH

_______________________________________________________
6.1 Individual contribution

Sl. No Topics Name

I Introduction to the Project MANOJ K
II Identifying the Problem YESHWANTH
III Objectives of the Thesis YESHWANTH
IV Process/Methodology TRINATH REDDY

V Literature Survey TRINATH REDDY

19
Chapter 7
CONCLUSION AND FUTURE WORK

________________________________________________________
In conclusion, developing a user-centric mobile application for the identification of plant disease is an
essential first step toward fulfilling the stated client demands. The comprehensive solution combines state-
of-the-art picture recognition algorithms, augmented reality capabilities, and an aesthetically pleasing
interface to ensure accurate and captivating plant identification experiences. The program aims to bridge
the gap between what is currently available and what users need by prioritizing user preferences and
convenience through voice command features and customized profiles.

Future Work:

Future research could go in a number of ways to expand the scope and effectiveness of plant identification
methods as the field plant disease continues to grow. One potential research direction is to integrate state-
of-the-art technologies, such as artificial intelligence and machine learning, to continuously improve the
accuracy and efficacy of plant identification algorithms. This may involve using deep learning techniques
to manage a larger range of plant variants and minute features while ensuring the system remains
adaptable to new species and environmental conditions.

20
In the future, the application's database may be expanded to include additional medicinal plants, including
ones from different ecosystems and regions. This would need ongoing collaboration with botanical
experts, researchers, and communities worldwide to ensure a comprehensive and diverse knowledge base.
Furthermore, adding ethnobotanical information about how other cultures have traditionally used plants
can help users better understand the cultural significance of medicinal plants and add a valuable layer to
the application.Updates and improvements to applications should be guided by ongoing user feedback to
improve the user experience. Analytics pertaining to user involvement have the potential to disclose
preferences, issues, and areas for development. Future iterations of the program might look into
gamification elements to make learning more enjoyable and interactive, which would promote ongoing
user participation.

In other words, future plant disease will necessitate a continuous focus on technological advancement,
knowledge base expansion, user experience enhancement, and partnerships that support a more informed,
sustainable, and culturally enriching relationship with medicinal plants.

REFERENCES

[1] Herdiyeni, Y., Nurfadhilah, E., Zuhud, E. A., Damayanti, E. K., Arai, K., & Okumura, H. (2013). A computer-aided
system for tropical leaf medicinal plant identification. International Journal on Advanced Science, Engineering and
Information Technology, 3(1), 23-27.

[2] Putri, Y. A., Djamal, E. C., & Ilyas, R. (2021, March). Identification of Medicinal Plants Leaves Using Convolutional
Neural Network. In Journal of Physics: Conference Series (Vol. 1845, No. 1, p. 012026). IOP Publishing

[3] H. X. Kan, L. Jin, and F. L. Zhou,” Classification of Medicinal Plant Leaf Image Based on Multi-Feature Extraction”,
Pattern Recognition and Image Analysis, Vol. 27, No. 3, 2017, pp. 581– 587, 1054-6618. © Pleiades Publishing, Ltd.

[4] Riddhi H. Shaparia, Dr. Narendra M. Patel and Prof. Zankhana H. Shah,” Flower Classification using Texture and
Color Features”, International Conference on Research and Innovations in Science, Engineering &Technology,
Volume 2, 2017, Pages 113–118

[5] Pushpa BR, Anand C and Mithun Nambiar P,” Ayurvedic Plant Species Recognition using Statistical Parameters
on Leaf Images”, International Journal of Applied Engineering Research, Volume 11, Number 7, 2016, 0973-4562

[6] Sandeep Kumar E,”Leaf color, area and edge features-based approach for identification of Indian medicinal
plants”, Indian Journal of Computer Science and Engineering (IJCSE), Vol. 3 No.3 Jun-Jul 2012, 0976-5166.

21
[7] Nawkar, M.G.; Maibam, P.; Park, J.H.; Sahi, V.P.; Lee, S.Y.; Kang, C.H. UV-induced cell death in plants. Int. J. Mol.
Sci. 2013, 14, 1608–1628.

[8] Salehi, B.; Kumar, N.V.; ¸Sener, B.; Sharifi-Rad, M.; Kılıç, M.; Mahady, G.B.; Vlaisavljevic, S.; Iriti, M.; Kobarfard, F.;
Setzer, W.N.; et al. Medicinal plants used in the treatment of human immunodeficiency virus. Int. J. Mol. Sci. 2018,
19, 1459.

[9] Pferschy-Wenzig, E.M.; Bauer, R. The relevance of pharmacognosy in pharmacological research on herbal
medicinal products. Epilepsy Behav. 2015, 52, 344–362

PLAGIARISM REPORT

22
23
24