Economy gray as plants approach maturity; nodes turn

Economy highly depends on
agricultural productivity.
Increasing amount the growth of crops need automatic monitoring disease. Detecting
disease from the images of the rice crops is one of the interesting research
areas. This survey presents different image processing techniques used in
detection of rice crops images of different techniques but also discuss
concepts of image processing applied to rice plant disease detection and
classification. Its include size of image dataset no. of diseases, segmentation
techniques, pre-processing, accuracy of classifiers etc. We utilize our survey
to design our work on detection and classification of rice plant diseases.

Image processing, classification; clustering, disease classification, disease


Farming is the most common source of the income
in indian people. Rice is the most cultivated food all over the world. The
losses of crops it brings indian economy decrease on agricultural field because
70% of the indian population depends on producing crops. Rice disease destroys
10 to 15% of production in Asia. Fungus, bacteria and viruses are responsible
for disease in the plant, so monitoring of disease on plant an important role
in the successful cultivation. Different disease that occur on rice plants are
leaf blast, brown spot, sheath blight and leaf scald. This survey
focuses on how image processing is utilized in detection of diseases in rice
plants. Disease identification. The rice crop diseases are discussed in detail.


1- Leaf Blast Disease: A region varying from
small round, dark spot to oval spots with narrow reddish-brown margins and gray
or white centre.


2- Brown Spots Disease: Round to oval shape with dark brown lesions. its occurs on leaves of the rice    plant.


3- Bacterial Blight Disease: Lesions consist of elongated lesions near the
leaf tip. turn white to yellow and then gray due to saprophytic fungi.


4- Sheath Blight Disease: Lesions consist of alternating wide band of
white, reddish-brown or brown. Fungal survival structures called sclerotia may
from on the leaf surface. under favourable conditions, bird nest area of dead
tissue may form.


5- Sheath Rot Disease: General reddish-brown discoloration of flag leaf
sheath, panicles emerging poorly; white frosting of conidia on inside of leaf
sheath, florets discolored a uniform reddish-brown or dark brown.


6- Node Blast Disease: Clum mode turns black and gray as plants approach
maturity; nodes turn dark to blue-gray.




Leaf blast


Brown Leaf Spot


Bacterial Blight


Sheath Blight


Sheath Rot


Node Blast


          Fig. 1. Different Types
of Rice Leaf Diseases



Here we describe different works that
already done by researchers in different fields such as leaf disease
classification, classification and segmentation of rice leaf. so aim of the
survey used all these control methods for a good harvesting rice plantation.

Suraksha I.S., et al,” Disease Prediction of
Paddy Crops Using Data Mining and Image Processing Techniques, IJAREEIE, Vol.
5, Issue 5, ISSN: 2320-3765  (2011).

First, the input is digital a colour image of paddy
disease leaf. Then a method of mathematics morphology is used to segment these
images. Erosion method has been used to removes small-scale details from a
binary images but simultaneously reduces the size of regions of interest. The
dilation is one of the basic operations in mathematical morphology. The
dilation operation usually uses a mesh for expanding the shapes contained in
the input image. 


Santanu Phadikar, et al, ” Rice diseases
classification using feature selection and rule generation techniques”
Computers and Electronics in Agriculture (2012).

Proposed a method classifying diseases of
the rice plant. In their approach, fermi energy based region extraction method
is applied to overcome the limitation of selecting the proper threshold value.
To identify the shape of the infected region, GA is applied that best
approximates the structure of the region. The position of infection is
determined by partitioning the spot into different blocks and arranged as a
quadtree at different lables. The binary representation of each block reduces
computational complexity reasonably. Using rough set concept features are
selected by generating all reduces which minimize loss of information. From the
reduced dataset a set of classification rules is derived using a novel
classification rule mining technique. The advantage of the proposed method is
that it does not require any gain calculation of the rules and so involves
lesser computational complexity.


S. Phadikar, et al,” Classification of Rice Leaf
Diseases Based on Morphological Changes,” International Journal of
Information and Electronics Engineering (2012).


Cultivation the earliest
and accurate diagnosis of the rice plant diseases able to reduce the damage, resulting
environment protection. In this survey, an automated system has been developed
to classify the leaf brown spot and the leaf blast diseases of rice plant based
on the morphological changes of the plants caused by the diseases. Bayes’ and
SVM Classifier have used radial distribution from the centre to the boundary of
the spot images as features to classify the diseases. The system has been
validated using 1000 test spot images of infected rice leaves collected from
the field, gives 79.5% and 68.1% accuracies for Bayes’ and SVM Classifier based
system respectively. In this survey, an automated system has been developed for
identifying two different types of rice diseases. In the first stage,
uninfected and the diseased leaves are classified based on the number of peaks
in the histogram. Miss classification may occur due to shadow effect and colour
distortion of aging leaves. In the second level, Bayes’ classifier and SVM are
applied to classify the leaf diseases 9. Time complexity of the
Bayes’ classifier is O (N×D2) where as for the support vector machine it is O (D×N2)
where D is the dimension of the feature vector and N is the number of training
samples. Since number of samples normally much larger than the dimension of the
feature vector, therefore the system is time efficient compare to SVM 2.