Tomato Leaf Disease Detection Using Deep Learning

Sreya Chowdary Karuturi

, Vinnakota Sai Vivek

, Charan Mandava

, Suthapalli Rishik

Tripty Singh

, Jyotsna C

and Prakash Duraisamy

Dept. of Computer Science and Engineering, Amrita School of Computing, Amrita Vishwa Vidyapeetham, Bengaluru,

India

Computer Science Department, University of South Alabama, U.S.A.

Keywords: Plant Leaf Disease Detection, Computer Vision, Image Classification, Crop Disease Detection, Convolutional

Neural Networks, Fully Connected Networks, MLP, Plant Leaves

Abstract: The recognition of crop diseases is essential to increase the yield and reduce quantity losses in agricultural by

products. Plant diseases pose a significant threat to global food security, yet early detection remains

challenging in many parts of the world due to the absence of essential infrastructure. Almost 60 percent of the

population is related to any kind of agriculture. The combination of expanding smartphone usage and the

breakthrough in computer vision and Image classification driven by deep learning has prepared the path for

smartphone or cellphone assisted-illness detection. This research presents a deep learning framework

identifying the disease on tomato leaves. By utilizing convolutional neural networks (CNNs), the system

extracts distinguishing features from input images and classifies them into one of the multiple classes of

diseases. The dataset used for training and evaluation consists of diverse images of different tomato plant

leaves with various diseases, collected from different sources. Experimental results demonstrate the model's

capability to achieve a 99.38% accuracy on test data, outperforming existing approaches. The proposed

approach has the potential to assist farmers and researchers in monitoring the health of plants, enabling them

to take timely action to prevent or minimize yield losses due to diseases.

1 INTRODUCTION

India has huge amount of population depending on

agriculture. Farmers have different options for

picking various acceptable crops and finding

appropriate insecticides for plants. Plant disease

reduces quality and quantity of farm products

s i g n i f i c a n t l y w h i c h m a y a l s o l e a d t o d a m a g e t o t h e

consumers food health. Crop disease research is

mainly focused on the examination of visually seen

patterns on plants. Monitoring plant diseases is

critical for an effective crop production in the farm or

for an agriculture industry. In the previous times both

plant monitoring, and analysis of it was done

manually by a knowledgeable person in agriculture.

I t r e q u i r e s a s i g n i f i c a n t l o t o f l a b o u r a s w e l l a s a

lengthy processing time. However, to identify plant

diseases we can use Image processing technique. This

study supports more in focusing on the visibly

intended crop quality.

Figure 1: Sample of Tomato disease identification and it’s

process in deep learning for the identification of the disease.

In most situations, illness signs may be found on

the leaves, stems, and fruit. The plant leaf is

recommended for disease detection since it displays

disease signs. Traditionally recognizing plant leaves

disease is a costly process as everything is done by an

expert, and hence inadequate for precision agriculture

since they need manual examination. To replace these

untrustworthy methods of disease diagnosis, research

has looked at image processing techniques on plant

photos. This study provides an outline of image

processing techniques used for plant disease

identification. Several conventional machine learning

Karuturi, S. C., Vivek, V. S., Mandava, C., Rishik, S., Singh, T., C, J. and Duraisamy, P.

Tomato Leaf Disease Detection Using Deep Learning.

DOI: 10.5220/0013639900004664

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 3rd International Conference on Futuristic Technology (INCOFT 2025) - Volume 3, pages 675-682

ISBN: 978-989-758-763-4

675

methods can be employed for detecting diseases in

plant leaves, including Support vector machines

known as SVM, random forests. To diagnose the

condition and determine its severity, machine

learning techniques were also applied.

Traditional machine learning approaches, on the

other hand, still depended on extracting crucial

characteristics from inputs for training models will be

a time- consuming process. Moreover, classical

image processing and machine learning were only

effective under certain situations. As a result, research

in the recent decade began to combine deep learning

algorithms that automatically perform feature

extraction and produce greater accuracy than

previous approaches with less time.

Deep learning can automatically extract

characteristics from photographs, it is one of its

advantages. Neural networks acquire a knowledge of

how to extract characteristics in the preparation time.

Convolutional Neural Networks, a type of multi-layer

feedforward neural network, are the most well-known

deep learning models. CNNs perform clustering by

grouping a vast number of image pixels into a smaller

number of clusters. However, determining the

optimal number of clusters can be challenging, as

different cluster sizes can lead to varying types of

image segmentation.

For the validation and classification of various

plant diseases from leaf images, several partitioning

and machine learning techniques have been proposed.

These methods have helped address existing

challenges, but the next step is to effectively

communicate the findings. Enhancing yield

efficiency in this sector remains a key objective.

2 LITERATURE SURVEY

Many researchers have focused on developing

detection of plant leaf disease. The highest accuracy

generated till now for plant leaf disease detection is

99.35%. Plant is one of the most widely used crops

the world by providing good nutrition and proteins to

our health. plants are affected by various diseases like

fungal, viral, and bacterial diseases which in turn

affect the plant growth and yield. The disease caused

by plants is harming our health. For good health,

yield, and plant growth we need to minimize plant

disease. To do that we have to predict the detection of

plant disease early.

Tiwari, 2020 explores the detection of potato leaf

diseases using deep learning models. The study

employs a Convolutional Neural Network on a

dataset of both healthy leaves and leaves affected

by four diseases. VGG16 architecture was utilized for

feature extraction, and classification is performed

with the use of Support Vector Machine.

Jiang, 2019 is about detection of apple leaf

diseases improved Convolutional Neural Networks

(CNN) in real-time. The authors proposed an

improved version of the CNN for disease detection in

apple leaves.

H. F, 2021 is about plant disease detection mobile

application development using deep learning. In this

study, the author utilized the Faster R-CNN with an

Inception-v2 backbone network for the application,

achieving an accuracy of 97.9%.

A. Lakshmanarao, 2021 is about plant disease

prediction and classification using deep learning

ConvNets. ConvNets were applied to three separate

datasets, achieving accuracies of 98.3% for potato

plants, 98.5% for pepper plants, and 95% for tomato

plants in disease detection.

S. Veni, 2021 the use of content-based image

retrieval for identifying plant leaves and detecting

diseases. They have used various image processing

techniques on images of leaves for recognising the

leaf plant type and for detecting the disease. Also, the

authors utilized two different classification

techniques. They are SVM, KNN and compared their

performance.

M. Kirola, 2022 presents a plant disease

prediction framework: an image-based system

utilizing deep learning. The authors employed various

machine learning (ML) and deep learning (DL)

algorithms, including Convolutional Neural

Networks (CNN) for disease prediction in plants.

They compared the performance of ML and DL

techniques, achieving a 97.12% accuracy with the

Random Forest classifier.

P. B, 2021 focuses on classifying plant diseases

using deep learning models. The authors utilized a

Convolutional Neural Network (CNN) based on the

AlexNet architecture. This model was compared to

other CNN models based on VGG-18 and LeNet-5

architectures. The study achieved accuracy of 96.76

with CNN model.

S. N, 2022 uses Convolutional Neural Network

(CNN) algorithm and linear regression analysis to

evaluate model performance. discovered that

increasing the number of images leads to higher

model accuracy, especially when images have clearer

visibility, compared to fewer images.

P. Sudharshan, 2022 utilizes a Support Vector

Machine (SVM) classifier to identify specific disease

affecting rice plants. The authors identified disorders

from their texture, shape and the colour of the rice.

INCOFT 2025 - International Conference on Futuristic Technology

676

R. G, 2018 employs the Faster R-CNN (F-RCNN)

detection model, achieving a confidence level of 80

and an overall accuracy of 95.75%. Also, the authors

evaluated the model's accuracy for tomato leaf

disease detection using automatic image capturing,

resulting in an accuracy of 91.67%.

P. K, 2022 presents a multi-layer deep learning

model for detecting potato leaf diseases. The model

achieved an impressive accuracy of 99.76%.

Shruthi, 2019 provides a review of machine

learning classification techniques for plant disease

detection. The authors highlight the use of

Convolutional Neural Networks (CNNs), noting their

high accuracy and ability to identify multiple diseases

in leaves.

K. Aparna, 2018 focuses on weed detection by

employing shape and size analysis. For detection of

the leaves that are affected they have used

thresholding methods.

T. Postadjian, 2018 discusses the effectiveness of

Convolutional Neural Network (CNN) architectures

in capturing disease-specific patterns from leaf

images. Additionally, the paper covers feature

extraction techniques such as transfer learning and

highlights evaluation metrics used to assess model

performance.

S. Ghosal, 2018 encompasses a wide range of

Convolutional Neural Network (CNN) architectures,

discussing their effectiveness in accurately

identifying plant diseases. The paper also explores

preprocessing techniques designed to enhance image

quality and discusses data augmentation strategies for

generating diverse training instances.

Y. Guo, 2020 explores various network

architectures, feature extraction methods, and image

augmentation techniques used for precise disease

classification. The paper discusses the efficacy of

these techniques in capturing and extracting relevant

features from leaf images.

D. Pujari, 2013 covers various architectures,

datasets, preprocessing techniques, and challenges in

the plant disease detection. It also discusses popular

models like Convolutional Neural Network (CNN)

and Recurrent Neural Network (RNN) and their

effectiveness in capturing relevant features. It

highlights the importance of diverse datasets and

explores preprocessing techniques for improving

image quality. The review also addresses challenges

such as class imbalance, limited labelled data, and

interpretability of model predictions.

H. Park, 2017 focuses specifically on deep

learning techniques for plant disease recognition. The

study provides a complete examination of different

network architectures, transfer learning methods, and

datasets. The survey highlights the importance of

datasets for training and evaluating models in the

field of plant disease recognition.

V. Pooja, 2017 employs digital image processing

techniques for detecting and classifying plants

disease using different classifier and different

Convolutional Neural Network (CNN) techniques

which improved the accuracy of the model overall.

N. K. E, 2020 utilizes ResNet-50 for pretraining

their model, with the implementation developed in

PyTorch. The study focuses on six types of diseases

and achieved an accuracy of 97%.

Below are the few literature survey questions by

that we will review the current state of research on

plant leaf disease detection.

I. What are the factors that are affecting the

growth of plant leaves and their yield?

 Soil quality

 Temperature and light exposure

 Pesticides

 Watering technique

II. What are the most common and important

pests that are causing damage to the tomato

leaves and should be identified early?

 Early and late blight are most common

diseases which are affecting tomatoes.

 Spider mites, budworm, two-spotted

mite, thrips, and caterpillars are the

most common pests affecting tomatoes.

III. What is the nutritional content of tomatoes

when consumed by humans?



The healthy tomato consists of vitamin

C, potassium and calcium and many

other needful nutrients for the body.

 Over consumption of tomatoes may

result in heartburn and acid reflux

which can get even worse if consumed

over the limit.

The novelty of this deep learning work is used in

search of different self-operated detection techniques

of tomato diseases and provide a user-friendly

interface and accurate detections rate to the farmers

to find the disease in field. The inspection of studies

shows the historic progress made in evolution of

detection systems based on deep learning, machine

learning, hyperspectral imaging, smartphone-based

technologies, and cost efficiency detection systems.

These systems provide several advantages, including

high precision, real-time detection, feasible, and cost

efficient. Moreover, these systems have the capability

Tomato Leaf Disease Detection Using Deep Learning

677

to decrease the spread of diseases, and enlarge crop

yields, which gives a better output in improved food

security and prosperity.

The use of automated detection for tomato leaf

diseases detection aligns with several UN SDGs. By

detecting diseases early and accurately, these systems

can minimize crop damage and increase yields, thus

contributing to SDG 2. The early detection and

management of diseases also promotes plant health,

ensuring healthy crops and contributing to SDG 3.

3 DATASET DESCRIPTION

The dataset we used have 11600 images of 5 different

classes which are Early Blight leaves, Healthy leaves,

Late Blight leaves, Leaf Mold leaves and Bacterial

Spot leaves. The dimension of every images is

256x256 pixels.

4 PROPOSED SYSTEM

4.1 Preprocessing & Data

Augmentation

The collected information known as images may have

outlier and unclear images, so we must preprocess

them or clean them to build up a model.

Comparatively preprocessed or cleaned data will give

better accuracy than uncleaned data. Resizing

should be done to constant size as the images come

from various locations, we must resize the images to

maintain consistency. Normalizing the pixel value

will make the neural network to learn easily. Data

augmentation can also be done it does flip, shifting

and changing the brightness on the images. Blurring

of image is done to reduce the noise of the image.

4.2 Study of MLP

Disease classification using CNNs was compared

to classifications using SVM and MLP algorithms to

determine the best method for diagnosing tea leaf

illnesses from photos. The latter two classifiers'

image features were derived by utilizing the bag of

visual words (BOVW) model, which is based on the

dense scale-invariant feature transform (DSIFT).

Disease Dataset: Using a Cannon camera, pictures

of tea leaf diseases were all taken in Yichang, Hubei

province, China, in their natural settings. The

photographs had a resolution of 4000 3000 pixels and

were taken in auto-focus at around 20cm above the

leaves.

The symptoms of seven different diseases, as

determined by phytopathologists, were depicted on a

total of 3810 photographs. Every image included in

the current manuscript was reduced in size to 256 x

256 pixels. We increased the dataset size, which is

better for the network's training, to enhance the

classifier's generalization capabilities.

4.3 Network Architecture of MLP

Figure 2: Network architecture of MLP.

The whole dataset is trained and tested using the

Multi-Layer Perceptron (MLP) Classifier. The

network architecture has six stages as shown below.

Classification using MLP network is a common

technique used in leaf disease detection. The basic

idea is to train a neural network on a set of labelled

leaf images (i.e. images with known disease labels) to

learn to recognize patterns that distinguish between

healthy leaves and with different diseases.

INCOFT 2025 - International Conference on Futuristic Technology

678

4.4 Study of Convolutional Neural

Networks

Convolutional Neural Networks (CNNs) have

become widely adopted for detecting leaf diseases,

representing a significant application of machine

learning and deep learning in agriculture. CNNs

possess the inherent capability to autonomously

extract relevant features from images through their

layered convolutional filters.

Firstly, we gathered high-resolution images of

tomato leaves, capturing both healthy specimens and

those exhibiting signs of disease. The collected

images effectively depict various diseases that impact

tomato plants. The images underwent resizing and

preprocessing to optimize them for CNN analysis.

The CNN model was trained using this prepared

dataset, incorporating convolutional, pooling, and

fully connected layers. Once trained, the model was

tested on a distinct validation dataset to determine its

effectiveness in recognizing diseases in plant leaves.

After development and testing, the model is ready

for practical use in detecting diseases in plant leaves

and it has been highly effective. We have used

because it can learn and extract relevant features from

images has led to improved accuracy in disease

detection compared to traditional methods.

Moreover, it can help farmers to identify diseased

plants early, preventing the spread of disease and

minimizing crop loss.

4.4.1 Architecture of Convolutional Neural

Network

The proposed Convolutional Neural Network (CNN)

architecture is tailored for accurate detection and

classification of tomato diseases. We normalized

input images of size 256 × 256 pixels for efficient

training. In our model first convolutional layer uses

24 filters with an 11 × 11 pixel kernel and a stride of

2, producing 24 feature maps of size 55 × 55 pixels.

ReLU activation introduces non-linearity, followed

by a 3 × 3 pooling layer and Local Response

Normalization (LRN), reducing the feature map size

to 24 × 27 × 27 pixels. The second convolutional

layer, with 64 filters and a 5 × 5 kernel, outputs 64

feature maps of size 27 × 27 pixels. Batch

Normalization and pooling layers further reduce the

size to 64 × 13 × 13 pixels. Subsequent layers utilize

96 filters with 3 × 3 kernels, producing 96 feature

maps of size 13 × 13 pixels. The final convolutional

layer uses 64 filters with the same kernel size.

Figure 3: Architecture of Convolutional Neural Network.

Flattened feature maps are passed to a fully

connected layer that maps feature to five disease

categories: Early Blight, Healthy, Late Blight, Leaf

Mold, and Bacterial Spot. A SoftMax layer performs

classification. Regularization techniques, including

dropout and the Adam optimizer, ensure robust

training and prevent overfitting, enabling the model

to achieve high classification accuracy.

4.4.2 Training and Validation Plots

The model has been trained on the training dataset of

plant leaf images and training accuracy is calculated

for it. As the number of training epochs increases, the

training accuracy increased as well, as the model

become more familiar with the training data and

better at predicting its class labels. During the

validation phase, we evaluated the model on a

separate dataset called the validation dataset. This

validation accuracy is calculated and plotted against

the training accuracy.

Figure 4: Validation Loss.

By analyzing the training and validation accuracy

plot, you can get a better understanding of how well

deep learning model is performing and adjust as

necessary. By analyzing the training and validation

loss plot, you can get a better understanding of how

well your deep learning model is performing and

adjust as necessary. For example, if the training loss

is decreasing but the validation loss is increasing, it

may indicate that the model is overfitting, and you

may need to apply techniques such as regularization

or early stopping to prevent overfitting.

Tomato Leaf Disease Detection Using Deep Learning

679

Figure 5: Validation Accuracy.

4.4.3 Classifiers

The classifiers we used are MLP, Support Vector

Machine also known as SVM, Random Forest,

XGBoost and CNN. In all we got CNN with highest

accuracy compared to other classifier models.

4.4.4 Regularization Study

Regularization techniques are used in deep learning

to prevent overfitting of the model, which occurs

whenever you have an issue like your model is

performing well on test data but not on the training

data. Then we will use regularization techniques. In

deep learning models, overfitting can be a common

issue, especially when the model is complex and

contains a lot of parameters.

A penalty term is added to the loss function using

L1 and L2 regularization approaches, which compels

the model to have light weights. As a result, the

model's complexity is decreased, which give lower

overfitting. Another method, called dropout

regularization, reduces the dependency of neurons

during training to reduce overfitting by randomly

removing certain neurons. Early stopping is applied

to terminate training when the model's performance

on the validation set ceases to improve, either after a

model set number of epochs or when additional

epochs yield no further gains.

The study demonstrates that incorporating L1 and

L2 regularization enhances the CNN model's

performance by mitigating overfitting. Dropout

regularization lessens overfitting, which enhances the

performance of the model. Similarly, we have used

early stopping to prevent overfitting and to end the

training process before the model begins to overfit.

Overall, the regularization methods we have used

in the study have improved the CNN model's

performance on the dataset with encouraging

findings. To get the best results, it is advised to

combine various strategies, as some of them might be

more effective for models and datasets. For deep

learning practitioners, regularization techniques are a

crucial tool to avoid overfitting and boost model

performance.

4.4.5 Optimizers

By using optimization, the loss will be reduced. It is

done by adjusting our weights and bias of the trained

model. Our training model is trained with image

samples similarly it is tested with testing image

samples it is done to compare the performance of both

the models. We used two optimization techniques as

they gave the better results, they are Adam’s

optimization technique and stochastic gradient

technique with k iterations and momentum. By using

optimization, the loss will be reduced. It is done by

adjusting our weights and bias of the trained model.

As we have many neural networks in this deep

learning model optimization techniques are needed to

use. The benefit we get by using optimization

techniques is that we get a good performance by

decreasing the loss function of the training model. In

our project all optimizers have achieved almost the

same validation and testing accuracy, which is 0.8 so

that is a good score for our training techniques. The

all-different optimization techniques which we used

are stochastic gradient descent technique and Adam.

We got the best results for them but there are few

other optimization techniques we used which are

AdaDelta, RSMProp, SGD, and Adagrad.

Adam’s and Stochastic gradient descent

optimization techniques stood out for our training

model as they gave better results. Adam’s

optimization technique is the mix of two optimization

techniques which are SGD and RMSprop with

momentum. In the optimization technique to fit the

model to the best weight for the neural network it is

updated using backpropagation algorithm.

5 RESULTS

The dataset we used in this study includes leaf images

from various plants such as tomato, potato, pepper,

and weed. To ensure uniformity, the images

underwent preprocessing techniques like resizing and

normalization. Both CNN and MLP models were

tested, with CNN delivering the highest accuracy of

99.38%, significantly outperforming MLP, which

achieved 78%. This superior performance of our

model can be attributed to its ability to efficiently

capture spatial and visual patterns in images.

Regularization methods such as L1 (48%

accuracy), L2 (52% accuracy), and Dropout (45%

accuracy) were applied to reduce overfitting and

INCOFT 2025 - International Conference on Futuristic Technology

680

improve generalization. Additionally, several

optimizers were tested to enhance model training,

yielding the following accuracies: Adam (93.25%),

RMSProp (91.34%), SGD (87.45%), Adagrad

(89.32%), and Adadelta (74.25%). These

optimization techniques we used played a crucial role

in improving model convergence and stability.

In conclusion, We have created a user-friendly

interface with TKinter GUI which can be easily used

by farmers to capture images and the CNN model

proved to be the most effective for detecting plant leaf

diseases, achieving remarkable accuracy. By

integrating advanced regularization techniques and

optimizers, the study highlights the potential of deep

learning in agricultural applications.

6 FUTURE WORK

In the future, we aim to expand the dataset with more

diverse plant species and environmental conditions to

improve model robustness. Additionally, integrating

multi-disease detection capabilities and severity

classification will enhance its utility. Field testing

with feedback from agricultural experts will validate

the model, paving the way for practical

implementation in precision farming.

REFERENCES

Tiwari, D., Ashish, M., Gangwar, N., Sharma, A., Patel, S.,

& Bhardwaj, S. (2020, May). Potato leaf diseases detect

ion using deep learning. In 2020 4th International

Conference on Intelligent Computing and Control

Systems (ICICCS) (pp. 461-466). IEEE.

Jiang, P., Chen, Y., Liu, B., He, D., & Liang, C. (2019).

Real-time detection of apple leaf diseases using deep

learning approach based on improved convolutional

neural networks. IEEE Access, 7, 59069-59080.

H. F. Ng, C. -Y. Lin, J. H. Chuah, H. K. Tan and K. H.

Leung, "Plant Disease Detection Mobile Application

Development using Deep Learning," 2021 International

Conference on Computer & Information Sciences

(ICCOINS), Kuching, Malaysia, 2021, pp. 34-38, doi:

10.1109/ICCOINS49721.2021.9497190.

A. Lakshmanarao, M. R. Babu and T. S. R. Kiran, "Plant

Disease Prediction and classification using Deep

Learning ConvNets," 2021 International Conference on

Artificial Intelligence and Machine Vision (AIMV),

Gandhinagar, India, 2021, pp. 1-6, doi:

10.1109/AIMV53313.2021.9670918.

S. Veni, R. Anand, D. Mohan and P. Sreevidya, "Leaf

Recognition and Disease Detection using Content

based Image Retrieval," 2021 7th International

Conference on Advanced Computing and

Communication Systems (ICACCS), Coimbatore,

India, 2021, pp. 243-247, doi:

10.1109/ICACCS51430.2021.9441805.

M. Kirola, K. Joshi, S. Chaudhary, N. Singh, H. Anandaram

and A. Gupta, "Plants Diseases Prediction Framework:

A Image-Based System Using Deep Learning," 2022

IEEE World Conference on Applied Intelligence and

Computing (AIC), Sonbhadra, India, 2022, pp. 307-

313, doi: 10.1109/AIC55036.2022.9848899.

P. B R, A. Ashok and S. H. A V, "Plant Disease Detection

and Classification Using Deep Learning Model," 2021

Third International Conference on Inventive Research

in Computing Applications (ICIRCA), Coimbatore,

India, 2021, pp. 1285-1291, doi:

10.1109/ICIRCA51532.2021.9544729.

S. N, S. Nema, B. K. R, P. Seethapathy and K. Pant, "The

Plant Disease Detection Using CNN and Deep Learning

Techniques Merged with the Concepts of Machine

Learning," 2022 2nd International Conference on

Advance Computing and Innovative Technologies in

Engineering (ICACITE), Greater Noida, India, 2022,

pp. 1547-1551, doi:

10.1109/ICACITE53722.2022.9823921.

P. Sudharshan Duth and P. Rithik Lal, "Paddy Leaf Disease

Classification Using Machine Learning and Deep

Learning Techniques," 2022 International Conference

on Futuristic Technologies (INCOFT), Belgaum, India,

2022, pp. 1-6, doi:

10.1109/INCOFT55651.2022.10094429.

R. G. de Luna, E. P. Dadios and A. A. Bandala, "Automated

Image Capturing System for Deep Learning-based

Tomato Plant Leaf Disease Detection and

Recognition," TENCON 2018 - 2018 IEEE Region 10

Conference, Jeju, Korea (South), 2018, pp. 1414-1419,

doi: 10.1109/TENCON.2018.8650088.

P. K. Shukla and S. Sathiya, "Early Detection of Potato Leaf

Diseases using Convolutional Neural Network with

Web Application," 2022 IEEE World Conference on

Applied Intelligence and Computing (AIC), Sonbhadra,

India, 2022, pp. 277-282, doi:

10.1109/AIC55036.2022.9848975.

Shruthi, U., Nagaveni, V., & Raghavendra, B. K. (2019,

March). A Review on Machine Learning Classification

Techniques for Plant Disease Detection. In 2019 5th

International Conference on Advanced Computing &

Communication Systems (ICACCS) (pp. 281-284).

IEEE.

K. Aparna and P. Supriya, "Precision Agriculture in Maize

Fields," 2018 Second International Conference on

Intelligent Computing and Control Systems (ICICCS),

Madurai, India, 2018, pp. 1407-1410, doi:

10.1109/ICCONS.2018.8662936.

T. Postadjian, A. L. Bris, C. Mallet and H. Sahbi,

"Superpixel Partitioning of Very High Resolution

Satellite Images for Large-Scale Classification

Perspectives with Deep Convolutional Neural

Networks," IGARSS 2018 - 2018 IEEE International

Geoscience and Remote Sensing Symposium,

Valencia, Spain, 2018, pp. 1328-1331, doi:

10.1109/IGARSS.2018.8519222.

Tomato Leaf Disease Detection Using Deep Learning

681

S. Ghosal, D. Blystone, A. K. Singh, B.

Ganapathysubramanian, A. Singh, and S. Sarkar, “An

explainable deep machine vision framework for plant

stress phenotyping,” Proceedings of the National

Academy of Sciences, vol. 115, no. 18, pp. 4613–4618,

2018.

Y. Guo, X. Hu, Y. Zou et al., “Maximizing E-tailers’ sales

volume through the shipping-fee discount and product

recommendation system,” Discrete Dynamics in Nature

and Society, vol. 2020, pp. 1–14, 2020.

D. Pujari, R. Yakkundimath, and A. S. Byadgi, “Grading

and classification of anthracnose fungal disease of fruits

based on statistical texture features,”

International Journal of Advanced Science and

Technology, vol. 52, pp. 121–132, 2013.

H. Park, J. S. Eun and S. H. Kim, Image-based disease

diagnosing and predicting of the crops through the deep

learning mechanism, In Information and

Communication Technology Convergence (ICTC),

IEEE 2017.

V. Pooja, R. Das and V. Kanchana, "Identification of plant

leaf diseases using image processing techniques," 2017

IEEE Technological Innovations in ICT for Agriculture

and Rural Development (TIAR), Chennai, India, 2017,

pp. 130-133, doi: 10.1109/TIAR.2017.8273700.

N. K. E., K. M., P. P., A. R. and V. S., "Tomato Leaf

Disease Detection using Convolutional Neural Network

with Data Augmentation," 2020 5th International

Conference on Communication and Electronics

Systems (ICCES), Coimbatore, India, 2020, pp. 1125-

1132, doi: 10.1109/ICCES48766.2020.9138030.

INCOFT 2025 - International Conference on Futuristic Technology

682