International Journal of Innovative Technology and Exploring Engineering (IJITEE)

ISSN: 2278-3075, Volume-8 Issue-9, July, 2019

Smart Farming: IoT Based Plant Leaf Disease

Detection and Prediction using Deep Neural
Network with Image Processing
Prema K, Carmel Mary Belinda

 To monitor crops, various sensors, camera and IT unit are

Abstract: Agriculture plays a major role in human life. Almost used to capture images at regular intervals and then
60% of the population is involved directly or indirectly in some integrated into imaging system (uses advanced Machine
agriculture activity. But Nowadays, farmers have quit agriculture learning and Artificial Intelligence techniques) to get better
and shifted to other sectors due to less adoption of automation
yield and reduces crop failure. In future, machines
and other reasons like increase in the requirement of agricultural
laborers. So, Farmers now largely depend on adoption of integrated with Internet of Things and Image processing
cognitive solutions with technological advancements to acquire could entirely replace the need for humans to manually
the benefits. Image processing and Internet of Things jointly weed or monitor crops [5].
produces new dimensions in the field of smart precision farming.
This proposed methodology aims to create an approach for plant II. RELATED WORK
leaf disease detection based on deep neural network. This
approach combines IoT and image processing which runs pre- In past years, various traditional machine learning
processing and feature extraction techniques by considering algorithms have been used to disease diagnosis and analysis.
different features such as color, texture, size and performs Recently, several researchers have studied plant leaf disease
classification using deep learning model that expands to help
identification of plant leaf disease.
detection and classification with neural network algorithms
and image processing. Also, the research studies showed
Index Terms: Deep Convolutional neural network, Image that convolutional neural network have been used majorly
Processing, Internet of Things, Plant leaf disease detection. for plant disease recognition and showed good results.
The proposed approach combines the techniques of Internet
I. INTRODUCTION of Things and Image processing concepts to get expected
output.
In Agriculture field, Presence of Plant disease and
This paper mainly discusses on a conceptual framework that
organic deficiency causes substantial harm and loss to
certainly uses image processing model and deep learning
farmers. Traditionally, All the diseases and harms will be
neural network for classification.
identified with visual inspection by experienced people who
may use certain features like color, texture and shape to
analyze which in turn leads to expensive cost and less III. METHODOLOGY
efficiency. By considering this issue as a challenge, we As discussed earlier the Internet of Things system
aimed to provide a solution technically with combined includes sensors and cameras to capture the image of the
method of Internet of Things and Image Processing. plant leaf which can be divided as 80-20 ratio for training
Application of Internet of Things in agriculture includes and testing the images. The color, shape, texture and the leaf
agricultural monitoring and control, controlled environment size are the attributes used for diagnosis.
agriculture, open – field agriculture, livestock applications, a. Acquisition of Image
food supply chain tracking [3]. Benefits of smart farming In this conceptual approach, this stage indicates the input.
using Internet of Things improves agriculture in different The images are retrieved by using camera. The leaf image is
ways like tons of data collected by smart sensors, provides extracted and gets saved in the database for further process.
better control over the internal processes. Use of Internet of b. Image Pre-processing
Things in agriculture produces cost management, waste In this stage the noise removal and data normalization are
reduction, process automation and enhanced product quality used as pre-processing model in which all features are
and volumes. Precision agriculture mainly depends on normalized from vector to unit space.
image-based recognition. Automating this analysis is c. Image Analysis
especially beneficial for those a farmer to which expert In this step, segmentation of images is done to find the
knowledge and advice is not readily available or affordable. region of interest. In segmentation, the technique used is
Revised Manuscript Received on July 10, 2019 region-based segmentation which separates healthy and
Prema K, Computer Science and Engineering, VelTech Rangarajan
Dr.Sagunthala R&D Institute of Science and Technology, Avadi, Chennai,
diseased region of the plant leaf by using the color of the
TamilNadu, India. leaf [4].
Carmel Mary Belinda, Computer Science and Engineering, VelTech
Rangarajan Dr.Sagunthala R&D Institute of Science and Technology,
Avadi, Chennai, TamilNadu, India.

Retrieval Number I7707078919/19©BEIESP Published By:

DOI: 10.35940/ijitee.I7707.078919 Blue Eyes Intelligence Engineering
3081 & Sciences Publication
 Smart Farming: IoT Based Plant Leaf Disease Detection and Prediction using Deep Neural Network with Image
Processing
The features extracted from the image includes color, processing technique for monitoring and analysing the
texture and shape. Color feature contains information about conditions of healthy and diseased leaf.
boundary, spot and broken area. Likewise, the shape
attribute includes percentage of lesion and its type. Texture
feature contains uniformity, contrast, probability, variance
and correlation. With the identified features, the database is
divided into two sets of images for training and testing.

IV. BLOCK DIAGRAM OF PROPOSED APPROACH

Fig: - Test Picture

VII. CONCLUSION
The proposed approach is a conventional framework to
identify the plant leaf disease, which can discover the
features of leaf disease automatically. This approach turns
the traditional methods of farming to modern methods
which makes easy and economically productive. By
applying the techniques of Internet of Things bring
automation in the field of agriculture and when it gets
Fig: - Block Diagram integrated with image processing, the end user will find
satisfactory results. This in turn helps the end user to
V. CONVOLUTIONAL NEURAL NETWORK
monitor the environmental factors easily in spite of getting
Training images are used to build deep convolutional into field physically. Automating this approach helps the
neural net-work to extract the macro information about the farmers to monitor the field by working at their own places
image[1]. The structure of convolutional neural network without being present physically in the field. In future the
model contains convolutional layer, pooling layer, activation camera module can be placed as a drone to capture the fields
function and softmax layer. There are various CNN at different direction.
architectures which are available that can be used in this
approach concerning the identification of plant leaf disease REFERENCES
were: Alex Net, Google Net, VGG [2]. The training process 1. Fuentes, A., Yoon, S., Kim, S.C., Park, D.S., 2017, A robust deep-
of CNN includes optimization stage learning-based detector for real-time tomato plant diseases and pest
where the feed forward and back propagation pass will be recognition. Sensors 17, 2022, “.
2. Konstantinos P., Ferentinos, 2018, Deep learning models for plant
performed to minimize the error. disease detection and diagnosis. Computers and Electronics in
Agriculture.
3. Nurzaman Ahmed, Debashis De, Md. Iftekhar Hussain, 2018, Internet
of Things for smart precision agriculture and farming in rural areas
4. Sladojevic, S., Arsenovic, M., Anderia, A., Culibrk, D., Stefanovic, D.,
2016. Deep neural networks based recognition of plant diseases by leaf
image classification. Computational Intelligence neurosci.
5. Tichkule, S.K., Gawali, P.D.H., 2016. Plant disease detection using
image processing techniques.Green Engineering and Technologies.
6. Rastogi, A., Arora, R., Sharma, S., 2012. Leaf disease detection and
grading using computer vision technology & fuzzy logic. In: 2015 2nd
International Conference on Signal Processing and Integrated
Networks (SPIN). IEEE, pp. 500–505.
7. J Revathi, P., Hemalatha, M., 2012. Classification of cotton leaf spot
diseases using image processing edge detection techniques. In: 2012
International Conference on Emerging Trends in Science, Engineering
and Technology (INCOSET). IEEE, pp. 169–173.
8. Sannakki, S.S., Rajpurohit, V.S., Nargund, V., Kulkarni, P., 2013.
Diagnosis and classification of grape leaf diseases using neural
Fig: - Structure of Convolutional neural network networks. In: 2013 Fourth International Conference on Computing,
Communications and Networking Technologies

VI. RESULTS AND DISCUSSIONS

The objective of this proposed approach is to implement
modern technologies of image processing with deep neural
network model. Usage of Internet of Things system makes
the process easier which is extended to some level of
automation for capturing the images at regular intervals. The
captured image is used further for applying image

Retrieval Number I7707078919/19©BEIESP Published By:

DOI: 10.35940/ijitee.I7707.078919 Blue Eyes Intelligence Engineering
3082 & Sciences Publication
 International Journal of Innovative Technology and Exploring Engineering (IJITEE)
ISSN: 2278-3075, Volume-8 Issue-9, July, 2019

AUTHORS PROFILE

Mrs. K. Prema, Completed B.E., Computer Science

and Engineering in the year 2006. M.Tech., in the
discipline of Computer science and Engineering..
Currently working as Assistant Professor in Vel Tech
Rangarajan Dr.Sagunthala R&D Institute of Science and
Technology and pursuing Ph.D., in the area of image
processing and Internet of Things. Having 10 years of experience in
teaching field. Published research papers in image processing area.
Currently working on smart precision farming to improve green house
agriculture.

Dr. M. J. Carmel Mary Belinda, currently working as

Associate Professor in Vel Tech Rangarajan
Dr.Sagunthala R&D Institute of Science and
Technology.Completed her Ph.D. in the area Data
mining.Having 15+ years of experience in taching field.
Published 20 research papers in the field of Data Mining.

Retrieval Number I7707078919/19©BEIESP Published By:

DOI: 10.35940/ijitee.I7707.078919 Blue Eyes Intelligence Engineering
3083 & Sciences Publication