Volume 3, Issue 5, May – 2018 International Journal of Innovative Science and Research Technology

ISSN No:-2456-2165

Multiclass Classification Method Based on Deep

Learning for Leaf Identification for Farmers
Sanjay Sunkad Srujan Rumale
Dept of Computer Science and Engineering Dept of Computer Science and Engineering
Maharaja Institute of Technology Mysore Maharaja Institute of Technology Mysore
Mysore INDIA Mysore INDIA

Aakash Bagale Zabiulla Shariff

Dept of Computer Science and Engineering Dept of Computer Science and Engineering
Maharaja Institute of Technology Mysore Maharaja Institute of Technology Mysore
Mysore INDIA Mysore INDIA

Abstract:- There are estimated to be nearly half a million plants leave. Leaves are different from each other by
species of plant in the world. Classification of species has characteristics such as the shape, colour, texture and the
been historically problematic and often results in duplicate margin. Dataset are used for both training and testing.
identifications. Plant identification based on leaf is
becoming one of the most interesting and a popular trend. Neural networks are used here. Convolutional Neural
Each leaf carries unique information that can be used in Networks are very similar to ordinary Neural Networks they
the identification of plants. In the identification of plants are made up of neurons that have learnable weights and biases.
based on leaf, the leaf images need to be pre-processed Each neuron receives some inputs, performs a dot product and
accordingly to extract the various critical features. As the optionally follows it with a non-linearity. The whole network
upcoming ground-breaking performance of neural still expresses a single differentiable score function: from the
networks in the field of classification and identification of raw image pixels on one end to class scores at the other. And
objects is being carried out where success is being they still have a loss function (e.g. SVM/ Softmax) on the last
achieved, the new neural network known as Convolutional (fully-connected) layer and all the tips/tricks we developed for
Neural Network(CNN) has made the identification and learning regular Neural Networks still apply.
classification of objects more reliable with the
performance and as well as the computations. Like any
other classifier, the Convolutional Neural Network II. RELATED WORK
(Model) is trained with images and its specific labels.
Having few layered Neural Network; we will get patterns Talking about the existing system there are many
(features) in given object (image). This trained model is traditional machine learning methods like K-nearest neighbour
used to classify new input images. As supervised deep (KNN), Support vector machine (SVM) result in lower
learning is used it makes it produces accurate results and accuracy in prediction leading to miss classification of input
makes it easier for the user to classify leaf data with much image. Random forests, KNN, SVM have strengths in other
higher accuracy. CNN’s are great at images, and have type of problem: mainly in the space where there are relatively
promise for text, simply because they are capable of few (tens or hundreds) of heterogeneous features. CNN’s are
identifying patterns in huge number of homogeneous great at images, and have promise for text, simply because
features (pixels in 2D or characters in 1D). They are they are capable of identifying patterns in huge number of
designed to handle large amount of image data. Hence by homogeneous features (pixels in 2D or characters in 1D). They
using Convolutional Neural Network, the accuracy is are designed to handle large amount of image data. Hence by
drastically improved. using Convolutional Neural Network, the accuracy is
drastically improved. Traditional Image processing uses linear
classifier like support vector machine or K Nearest Neighbours
I. INTRODUCTION (KNN) where it fails to classify new data points into its
specific class. If such classifier is used to classify bulk images
Plant identification based on leaf is becoming one of the it could lead to reduction of their quality because of
most interesting and a popular trend. Each leaf carries unique misclassification.
information that can be used in the identification of plants. In
the identification of plants based on leaf, the leaf images need
to be pre-processed accordingly to extract the various critical
features.

The proposed approach consists of three phases that are

pre-processing, feature extraction, and classification phases.
Since most types of plants have unique leaves, so the
classification approach presented in this research depends on

IJISRT18MY274 www.ijisrt.com 563

 Volume 3, Issue 5, May – 2018 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
III. METHODOLOGY A. Equations
k-means Clustering

Fig 1:- Flow of process

A. Preprocessing
The input image is first converted into HSV color space
model and then using an inrange function which extracts Equation 1.
majority of the green pixels present in the image. Further a Back Propagation
bounding rect function is used to crop the image where the
unwanted region is eliminated. The cropped image that is
region of interest (ROI), which is given as input to the next
step for identification.

B. Feature Extraction and classification

CNN model is used for feature extraction where the data
is analyzed in the following manner where the CNN model
automatically extracts thousands of features. The image is
passed through a stack of convolutional layers. Response
normalization layers follow the first and second convolutional
layers. Spatial pooling is carried out by four max-pooling
layers. The ReLU non-linearity is applied to the output of
every convolutional and fully-connected layer.

Fig 2:- Inception V3 model

IV. ABBREVIATIONS AND ACRONYMS

HS - Hue Saturation
CNN - Convolutional Neural Network
ReLU - Rectified Linear Unit Fig 3:- Back Propagation

IJISRT18MY274 www.ijisrt.com 564

 Volume 3, Issue 5, May – 2018 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
 Module 2: leaf Recognition using Convolutional Neural
V. WORKING Network
Leaf recognition has always been a challenging task
from traditional image processing technique to modern image
processing. To obtain higher accuracy and better prediction of
the recognized Convolutional neural network is used where
Artificial Intelligence Model (CNN) needs to be integrated.
This involves the following steps. Collecting the dataset is an
important work where we need to collect images of leaves. If
we have 2 people, then we need a dataset of at least 20-25
images. More the data better the accuracy in modern AI.
Training the Model: We need to train the new leaves (20-25
images) and teach our model which the particular leaf is for
example, we load 25 images of leaf 1 and train the model and
20 images of leaf 2 and train the model.Leaf Recognition:
Finally, we should be able to detect the leaves that we have
been trained with good prediction accuracy.Retraining the
model: If the model has lower accuracy then we need to
retrain the model and start from step1 for better accuracy. We
need to make use of Tensor flow (Google’s Open Source
library) for face recognition where python is used as a base
language.

 Module 3: leaf disease recognition

The diseased leaf needs to be identified. The disease
may be categorized into many forms. We first need to identify
the disease of the leaf by giving the image of the leaf. Pre-
training the model is necessary as it was done in module 2.
The diseased leaf is categorized into its respective category.
Creating an UI using C# where an UI is integrated with some
buttons, picture boxes, timers as per coding requirement.
 Module 4: Connection leaf recognition code/leaf disease
recognition code with that of OpenCV C++ library. The
OpenCV C++ code needs to be connected with that of the
OpenCV code to make it simpler.
 Module 5: Display the final output on the UI
Final output has to be displayed on the output with no
glitch and error free.

VI. ACKNOWLEDGMENT

We would like to thank Sushma Koushik N (Asst Prof &

Guide), Deepu. R (Prof & HOD), Tejas P(Asst. Prof) from the
Department of Computer Science and Engineering, Maharaja
Institute of Technology Mysore for the support of our work.

REFERENCES

Fig 4:- Working [1] Siang Thye Hang*, Masaki Aono, “Open World Plant
Image Identification Based on Convolutional Neural
Network”, IEEE 2016.
 Module 1: Leaf capture using OpenCV [2] Ajla KULAGLIC and B.BerkÜSTÜNDA, “Neural
Leaf is captured using OpenCV. The image of the leaf Network with Wavelet Pre-processing for Wheat Growth
can be captured by using 3 methods Default camera: The Stage”, IEEE 2016.
laptop built in camera can be used to capture the image of the [3] Jobin Francis,AntoSahayaDhas D and Anoop B K,
individual. IP camera can make use of mobile device and “Identificationof leaf diseases in pepper plants using soft
capture HD images of leaf to get better accuracy and results. computing techniques”, IEEE 2016.
Web camera can also be accessed to get the image of leaf. [4] Yuanyuan Liu, dengwenzhou, fan tang, yipping meng and
weiming dong, “Flower classification via convolutional
neural networks”, IEEE 2016.

IJISRT18MY274 www.ijisrt.com 565

 Volume 3, Issue 5, May – 2018 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
[5] Adil Salman, Ashish Semwal, V. M Thakkar and Upendra
Bhatt Leaf Classification and Identification using Canny
Edge Detector and SVM Classifier”, IEEE 2016. [6]
Suhas S Athani and CH Tejeshwar, ”Support Vector
Machine-Based Classification Scheme of Maize Crop”,
IEEE 2017.
[7] Anjali Yadav, Malay KishoreDutta, Carlos M. Travieso ,
Jesús B.Alonso, “Automatic Identification of Botanical
Samples ofleaves using Computer Vision”, IEEE 2017.
[8] MonzurulIslam, AnhDinh, KhanWahid and Pankaj
Bhowmik,“Detection of Potato Diseases Using Image
Segmentation and Multiclass Support Vector”, IEEE
2017.
[9] Mukesh Kumar Tripathi, D. Y. Patil College of
Engineering, Ambedkar and Dr. Dhananjay D. Maktedar,
“Recent Machine Learning Based Approaches for Disease
Detection and Classification of Agricultural Product”,
IEEE 2017.
[10] Suchit Purohit, Savita Gandhi, Ronak Viroja and Naina
Chaudhary, “Automatic Plant Species Recognition
Technique using machine Learning approaches”, IEEE
2017.
[11] HonghuiXie,Zhun Fan,Wenji Li,YibiaoRong, Yang
Xiao, and Lei Zhao, “Tobacco Plant Recognizing and
Counting Based on SVM”, IEEE 2017.
[12]HalilDurmu,EceOlcayGüne,MürvetKrc, “Disease
Detection on the Leaves of the TomatoPlants by Using
Deep Learning”, IEEE 2017.

IJISRT18MY274 www.ijisrt.com 566