Leaf Disease Detection And Prevention Using Support Vector Machine

Based on Image Processing In Matlab

Aim:

Detecting the type of leaf disease present in the leaves of plants using an image
processing approach in Matlab is the main aim of this project.

Objective:

This project propose an image pattern classification to identify different disease in leaf
with a combination of texture and color feature extraction. The purpose of this research is to find
appropriate features that can identify leaf disease. Firstly, normal and diseased images are
collected and pre-processed. Then, features of shape, color and texture are extracted from these
images. After that, these images are classified by support vector machine classifier. A
combination of several features are used to evaluate the appropriate features to find distinctive
features for identification of type of disease. When a single feature is used, shape feature has the
lowest accuracy of and texture feature has the highest accuracy. A combination of texture and
color feature extraction results a highest classification accuracy. A combination of texture and
color feature extraction with polynomial kernel results best classification accuracy.

Introduction:

Agriculture is changing social and economic environment day today. Improper

management leads to loss in agricultural products [1]. Farmers lack the knowledge of disease and
hence they produce less production. Kisan call centers are available but do not service 24*7 and
sometimes communication will be fail. Farmers are not able to explain disease properly on call
need to analysis the image of affected area of disease. Images and videos of crops provide better
view and agro scientist can provide a better solution but it not getting to all farmers. The
experimental results indicate that the proposed approach is a valuable approach, which can
significantly support an accurate detection of leaf diseases in a less computational effort.
 In today’s modern digital world research are continuously trying to increasing the
collectively of plants [1]. They have archived by using developing the higher breed seeds and
plants. But one problem still exist which is a major concern of the cultivation of crop and that is
crop diseases and the pesticides problem. Due to these problems, the cultivation decreases and
hence all the farmers and in turn the country suffers the lack of cultivation of plant [2]. Many of
the time disease need to prevent at early stage, but it not happens then it damage the plants
[3][4]. Due to that whatever the investment need to do that also in loss, to avoiding all these
need to detect disease at early stages. Sugarcane is cultivated in long duration that is 10 to 18
months, that leads to attack of many diseases [5]. Fungi-caused diseases in sugarcane are the
most predominant disease that appears as spots on the leaves.

Existing Methods / Literature Survey

Earlier papers are describing to detect mainly pests like aphids, whiteflies, thrips, etc using
various approaches suggesting the various implementation ways as illustrated and discussed
below. A cognitive vision system that combines image processing, learning and knowledge-based
techniques. They only detect mature stage of white fly and count the number of flies on single
leaflet. They used 180 images as test dataset .among this images they tested 162 images and
each image having 0 to 5 whitefly pest. They calculate false negative rate (FNR) and false positive
rate (FPR) for test images with no whiteflies (class 1), at least one white fly (class 2) and for whole
test set. Extend implementation of the image processing algorithms and techniques to detect
pests in controlled environment like greenhouse. Three kinds of typical features including size,
morphological feature (shape of boundary), and color components were considered and
investigated to identify the three kinds of adult insects, whiteflies, aphids and thrips. Promote
early pest detection in green houses based on video analysis. Their goal was to define a decision
support system which handles a video camera data. They implemented algorithms for detection
of only two bio aggressors name as white flies and aphids. The system was able to detect low
infestation stages by detecting eggs of white flies thus analyzing behavior of white flies. Proposed
pest detection system including four steps name as color conversion, segmentation, reduction in
noise and counting whiteflies. A distinct algorithm name as relative difference in pixel intensities
(RDI) was proposed for detecting pest named as white fly affecting various leaves. The algorithm
not only works for greenhouse based crops but also agricultural based crops as well. The
algorithm was tested over 100 images of white fly pest with an accuracy of 96%. Proposed a new
method of pest detection and positioning based on binocular stereo to get the location
information of pest, which was used for guiding the robot to spray the pesticides automatically.
Introduced contextual parameter tuning for adaptive image segmentation, that allows to
efficiently tune algorithm parameters with respect to variations in leaf color and contrast.
Presents an automatic method for classification of the main agents that cause damages to
soybean leaflets,i.e., beetles and caterpillars using SVM classifier.

Proposed Method:

The proposed system is as shown in fig 1. It consist various stages including collection of
images of agricultural fruits for creation of database. Image segmentation is performed using
clustering techniques. Features of segmented images are stored in database with respective
image of agricultural fruits. Using support vector machine classifier we would be finding out type
of disease presents in image and give remedies to control it.

Shape feature extraction:

Shape feature extraction used in this project are solidity, extent, minor axis length and
eccentricity. This features taken from research in order to extract shape feature in diseased
region.

Eccentricity is used to recognize whether the rust shape is a circle or line segment.
Eccentricity is the ratio of the distance between the foci of the ellipse and its major axis length.
An ellipse whose eccentricity is 0 can recognized as a circle, while an ellipse whose eccentricity
can recognized as a line segment[9].

Minor axis length is used to measure length of axis of the diseased region. Minor axis
length is the length of the minor axis of the ellipse that has the same normalized second central
moments as the region (in pixels)[9].

Extent is used to measure area of diseased region that is divided by the area of the
bounding box. Extent is computed as the area divided by area of the bounding box[9].
 Solidity is used to measure area of diseased region divided by pixels in the convex hull.
Solidity is the proportion of the pixels in the convex hull that are also in the region. It is computed
by dividing the area by convex area[9].

Texture Feature Extraction

Gray Level Co-occurrence Matrix (GLCM) extract second order statistical texture
features[10]. Texture feature extraction used in this research are contrast, correlation, energy
and homogeneity. This features taken from research [3] to extract texture feature in leaf diseased
region.

Contrast of the pixel and its neighbors is calculated over all of the image pixels. Contrast
is used to measure contrast between neighborhood pixel.

Color Feature Extraction:

Color is a distinctive feature for image representation that is invariant with respect to
scaling, translation and rotation of an image[9]. Mean, skewness and kurtosis are used to
represent color as features. To do this, we transform RGB to LAB.

Support Vector Machine:

Training sample in support vector machine is separable by a hyperplane. This hyperplane

is computed according to the decision function, where w is a weight vector and b is a threshold
cut-off [3] . To maximize the margin, w􀗐 f and b have to be minimized.
 Figure 1: Flow Chart

There are 3 types of classifier are used to which classifier gives the better result. The back
propagation and feed forward classifiers are not detecting some pests in an image. But SVM gives
better result. SVM is a non-linear classifier, and is a newer trend in machine learning algorithm.

Advantages:

 Cheapest method to detect the type of disease and insects.

 Reduce the costs and amount of pesticide used for crops.

Application:

 Can be further implemented for mixture ratio of pesticide to be used in a field within a
particular day.
 The system can be further be added to spray the pesticides automatically based on
pesticides detected.
SYSTEM SPECIFICATIONS:

HARDWARE:

 System : Pentium IV 2.4 GHz.

 Hard Disk : 40 GB.
 Monitor : 15 VGA Colour.
 Mouse : Logitech.
 Ram : 512 Mb

SOFTWARE:

 Operating system : Windows XP/ Windows 7.

 Software Tool : Matlab 8.10.
 Coding Language : Matlab.
 Toolbox : Image processing toolbox.

References:

[1] K. Dang, H. Sun, J. P. Chanet, J. Garcia-Vidal, J. M. Barcelo-Ordinas, H.L. Shi and K.M. Hou,
”Wireless Multimedia Sensor Network for plant disease detections", France-China
International Workshop,18-20 Septembre2013, Clermont-Ferrand, France.
[2] R. Huddar, S. Gowri and S R Rupanagudi “Novel Algorithm for Segmentation and
Automatic Identification of Pests on Plants using Image Processing” Third International
Conference Computing Communication & Networking Technologies (ICCCNT), 26-28 July
2012.
[3] Manoj Mukherjee, Titan Pal and Debabrata Samanta, “Damaged Paddy leaf detection
using image processing", Volume 3, No. 10, October 2012.
[4] H. Al-Hiary, S. Bani-Ahmad, M. Reyalat, M. Braik and Z. ALRahamneh, "Fast and Accurate
Detection and Classification of Plant Diseases", International Journal of Computer
Applications (0975 – 8887) Volume 17– No.1, March 2011.
[5] Sanjay B. Patil, “Leaf Disease Severity Measurement Using Image Processing”,
International Journal of Engineering and Technology Vol.3 (5), 2011, 297-301.
[6] Vincent Martin and Sabine Moisan,” Early Pest Detection in Greenhouses”, Computer and
Electronics in Agriculture Journal, 2009.