LITERATURE SURVEY
Multiple numbers of methods are used by researchers and analysed the dieses of
the plants. The concept of this project is to identify the disease of the plant leaves
and obtain solution to that disease. Image processing techniques are used to
identify and arrange the plant leaf diseases. With the help of these methods, the
plant diseases can be identified in the beginning stage itself.
1. Saradhambal.G, Dhivya.R, Latha.S, R.Rajesh, “Plant Disease
Detection and Its Solution Using Image Classification”,
International Journal of Pure and Applied Mathematics, vol.119,
pp.879-884 No.14 2018.
 This project implements an innovative idea to identify the affected
crops and provide remedy measures to the agricultural industry.
 By the use of k-mean clustering algorithm, the infected region of the
leaf is segmented and analyzed.
 Feature extraction technique helps to extract the infected leaf and
also to classify the plant diseases.
2. Grape Leaf
Disease Identification
Using
Improved
Deep
Convolutional Neural Networks
 This paper has proposed a deep learning approach for the
identification of six common grape leaf diseases and healthy leaves.
 In view of the various sizes of grape leaf disease spots, Inception
structures were applied to the model for enhancing ability of the
multi-scale feature extraction.
 With the expanded data set, the proposed DICNN model was trained
to classify seven type grape leaves.
 According to the experimental results, the proposed algorithm realizes
a recognition accuracy of 97.22%, which gives better performance
than other popular transfer learning techniques.
 Compared with the standard ResNet and GoogLeNet architectures,
the proposed DICNN model realizes higher convergence speed during
the training process and higher accuracy.
3. International Journal of Innovative Technology and Exploring
Engineering (IJITEE), “Sooty Mould Mango Disease Identification
Using Deep Learning”, ISSN: 2278-3075, Volume-8 Issue-5S March,
2019 by Priyadharshini M. K, R. Sivakami, M.Janani.
 In this paper they have developed a Convolution Neural Network, for
the identification of plant disease identification using leaf images.
 The trained model required a low computational power to classify the
images, make it feasible to integration into the mobile application.
 The DL takes longer time to train but its testing is more efficient
compared to traditional method.
4. UDAY PRATAP SINGH 1 , SIDDHARTH SINGH CHOUHAN 2 , (Student
Member, IEEE), SUKIRTY JAIN 3 , AND SANJEEV JAIN 2 , (Member,
IEEE)”Multilayer Convolution Neural Network for the Classification
of Mango Leaves Infected by Anthracnose Disease.
5. Jayamala K. Patil1, Raj Kumar, “Advances in Image Processing for
Detection of Plant Diseases”, Journal of Advanced Bioinformatics
Applications and Research, vol.2, issue. 2, pp. 135-141, June-2011.
METHODOLOGY
Preprocessing and Training the model (CNN): The database is preprocessed such as
Image reshaping, resizing and conversion to an array form. Similar processing is also
done on the test image. A database consisting of about 32000 different plant
species is obtained, out of which any image can be used as a test image for the
software. The train database is used to train the model (CNN ) so that it can identify
the test image and the disease it has .CNN has different layers that are Dense,
Dropout, Activation, Flatten,Convolution2D, MaxPooling2D. After the model is
trained successfully, the software can identify the disease if the plant species is
contained in the database. After successful training and preprocessing, comparison
of the test image and trained model takes place to predict the disease.
Database collection: Initial step for any image processing based project is acquiring
proper database which is valid .Most of the time the standard database is preferred
but in certain circumstances we do not get proper database .So in such conditions
we can collect the images and can form our own database. The database is
accessed from crowdAI which is plant disease classification challenge. Data
available here is not labeled .So the first task is to clean and label the database.
There is a huge database so basically the images with better resolution and angle
are selected. After selection of images we should have deep knowledge about the
different leaves and the disease they have. Huge research is done from plant village
organization repository. Different types of plant images are studied and
corresponding. After detail study, labeling in done by segregating the images and
with different diseases