International Journal of Informatics and Communication Technology (IJ-ICT)

Vol.6, No.3, December 2017, pp. 209~217

ISSN: 2252-8776, DOI: 10.11591/ijict.v6i3.pp209-217  209

Density Based Clustering with Integrated One-Class SVM for

Noise Reduction

Md. Abdul Awal, Mohammad Jahangir Alam*, Md. Nurul Mustafa

Department of Computer Science and Information Technology, Southern University Bangladesh
739/A Mehidibag Road, Chittagong, Bangladesh.

Article Info ABSTRACT

Article history: The technology based modern agriculture industries are today‟s requirement
in every part of agriculture in Bangladesh. In this technology, the disease of
Received Aug 21st, 2017 plants is precisely controlled. Due to the variable atmospheric circumstances
Revised Oct 5th, 2017 these conditions sometimes the farmer doesn‟t know what type of disease on
Accepted Oct 19th, 2017 the plant and which type of medicine provide them to avoid diseases. This
research developed for crops diseases detection and to provides solution by
using image processing techniques. We have used Android Studio to develop
Keyword: the system. The crops diseases detection and solution system is compared the
image of affected crops with database of CDDASS (Crops Diseases
agricultural experts Detection and Solution system). If CDDASS detect any disease symptom,
android apps then provide suggestion so that farmers can take proper decision to provide
CDDASS medicine to the affected crops. The application has developed with user
Crops Diseases Detection friendly features so that farmers can use it easily.
Image processing
Copyright © 2017 Institute of Advanced Engineering and Science.
All rights reserved.

Corresponding Author:
Mohammad Jahangir Alam,
Department of Computer Science and Information Technology, Southern University Bangladesh
739/A Mehidibag Road, Chittagong, Bangladesh.
Email: [email protected]

1. INTRODUCTION
Agriculture and human social development go side by side as the production of crops made it
possible for primitive man to settle down in selected spots leading to formation of society [1]. The history of
agricultural in Bangladesh long before. Bangladesh is basically an agricultural country, and the income is
based on the agricultural products and all its resources depend on the agricultural output.
Although Bangladesh is on course for Middle Income Country status by 2021, agriculture remains
the largest employer in the country by far and 47.5% of the population is directly employed in agriculture and
around 70% depends on agriculture in one form or another for their livelihood. Agriculture is the source of
food for people through crops, livestock, fisheries; the source of raw materials for industry, of timber for
construction; and a generator of foreign exchange for the country through the export of agricultural
commodities, whether raw or processed. It is the motor of the development of the agro-industrial sector
including food processing, input production and marketing, and related services. As main source of economic
linkages in rural areas, it plays a fundamental role in reducing poverty which remains a predominantly rural
phenomenon [2].
Though, in Agricultural Department technology is rapidly changing, many automatic technologies
are coming in the market (example, Automatic planting, cutter machines etc which helps the farmer to
produce maximum products). Plant disease is an important concern for the farmers in Bangladesh because
Plant disease is an impairment of the normal state of the plant that interrupts or modifies its vital functions
[3]. To get solution on planet disease If farmers decide to take advice from agricultural expert regarding the
treatment of incidence of pest /disease/trait to their crop/plant in order to increase the crop productivity then
he may face following situations [4]:

Journal homepage: http://iaesjournal.com/online/index.php/IJICT

210  ISSN: 2252-8776

i) Sometimes they have to go long distances for approaching the expert.

ii) Even though they go such distances expert may not be available at that time.
iii) Sometimes, the expert whom a farmer contacts, may not be in a position to advise the farmer
with the available information and knowledge.
In these cases seeking the expert advice is very expensive and time consuming. Hence electronic
expert systems are needed. Electronic expert systems enable farmers in identifying type of diseases; making
the right decision and selecting the proper treatment. The expert systems are intelligent computer programs
that are capable of offering solutions or advices related to specific problems in given domain, both in a way
and at a level comparable to that of human expert in a field. One of the advantages of using Electronic expert
systems is its ability to reduce the information that human users need to process, reduce personnel costs and
increase throughput. Another advantage of expert system is that it performs tasks more consistently than
human experts [5].
There are various kinds of abnormality states present on the plants leaf which can be identified by
mean of manual inspection. The image processing and pattern recognition techniques play the worth full role
to convert manual process to automate the process. The automatic diagnosis system based on plant disease
features reduces the dependency on experts in the area concerned [6-8].Depending on the application, many
of those problems may be solved or at least reduced by the use of image processing.
Hence we are proposing an automatic crops diseases detection and solution system (CDDASS)
which helps a farmer to identify the disease of crops and provide treatment for his crops according to disease
using image processing techniques without help of any crops diseases expert. To get solution, farmers just
need to install the apps into their mobile phones for the first time and upload the affected crops‟ image to
detect and get suggestion from the system. That‟s way a farmer can save time, efforts and money and can be
tension free.

2. LITERATURE REVIEW
Prof.H.M. Deshmukh, Jadhav Sanjivani,Lohar Utkarsha, Bhagat Madhuri and Salunke Shubhangi.
“Plant Leaf Disease Identification System for Android”. International Journal of Advanced Research in
Computer and Communication Engineering [9]. In this article they have described the development of an
Android application that gives users or farmers the capability to identify the plant leaf diseases based on the
photographs of plant leaves taken through an android application.
Sufiyan K Shekh, Aniket Baitule, Milind Narethe, Sangappa Mallad, Waghdarikar and Dr D Y Patil.
“Detection of Leaf Diseases and Monitoring the Agricultural Resources using Android App”. International
Journal of Innovative Research in Computer and Communication Engineering [10]. In this article they have
described and designed a new architecture for remote control of agriculture devices and also detecting the
diseases of plant which make much easier and less dependent of the conditions present to farmers.
Mr.Sushant.S.Chavan, Mr.Nitesh.P.Satre, Mr.Rajat.R.Deshmane, Mr.Suhas.B.Katkar. „„Android
Based App to Prevent Crop Diseased in Various Seasons‟‟. International Research Journal of Engineering
and Technology(IRJET) [11]. In this article they have developed an android application for agriculture, when
and which fertilizers, pesticides and herbicides are used to save various crops from various diseases.

3. PROPOSED METHOD
3.1. Flow chart

Figure 1. Flowchart of Crops Diseases Detection and Solution System

IJECE Vol. 6, No. 3, December 2017 : 209– 217

IJ-ICT ISSN: 2252-8776  211

The flowchart describes the whole process of the research. At first user has to upload a defected
crops image on the system then system will search image related information in the system database, if match
the uploaded image with the system database it will provide possible diseases name with solution, if does not
match it will show not match message.

3.2. Interface Design

Figure 2. Home page of Crops Diseases Detection and Solution System

The figure shows main page of the system where users can choose their activities including menu
access and different crops list. A user can be able to see some diseases list with image.

Figure 3. Diseases List Page

Crops Diseases Detection and Solution System (Md. Abdul Awal)

212  ISSN: 2252-8776

Figure 4. Disease wise symptom and solution Page

After clicking any icon, image of diseases, visitor can be able to find specific disease symptoms
with solutions.

4. RESEARCH METHOD
There are some fundamental steps:
a. Image Acquisition: Images of the infected leaves have taken from the crops. The database has
different types images on plant diseases and the images are stored in (.jpg, .png) image format.
b. Image preprocessing: If noises are present in image, interested region in the image is not clear. In
the image clipping, smoothing, enhancement is the three steps included in preprocessing phase. The
process of image collection and lots of information may bring noise which make the quality of
image dropped. To perform demising different kinds of reduction technique are applicable.
c. Image Segmentation: Image segmentation is the first step and also one of the most critical tasks of
image analysis. According to the region of interest, the image will be segmented into different parts.
d. Image Feature Extraction: The features extraction is the input data transform into set of features.
The feature set will extract the relevant information. so should carefully have chosen. Feature
extraction involves simplifying the amount of resources required to describe a large set of data
accurately.
e. Image classification: The intent of the classification process is to categorize all pixels in a digital
image into one of classes or theme. The objective of image classification is to identify, as a unique
gray level (or color), features occurring in an image in terms of the object or type these features
actually represent on the ground. Image classification is perhaps the most important part of digital
image analysis [12].

5. BINARY ROBUST INVARIANT SCALABLE KEY POINTS (BRISK) : THE METHOD

Description of the key stages in BRISK, namely feature detection, descriptor composition and key
point matching to the level of detail that the motivated reader can understand and reproduce. It is important to
note that the modularity of the method allows the use of the BRISK detector in combination with any other
key point descriptor and vice versa, optimizing for the desired performance and the task at hand [13].

5.1. Scale Space Key Point Detection

With the aim of achieving invariance to scale which is crucial for high-quality key points, we go a
step further by searching for maxima not only in the image plane, but also in scale-space using the FAST
scores as a measure for saliency. Despite discretizing the scale axis at coarser intervals than in alternative
high-performance detectors, the BRISK detector estimates the true scale of each key point in the continuous

IJECE Vol. 6, No. 3, December 2017 : 209– 217

IJ-ICT ISSN: 2252-8776  213

scale-space. In the BRISK framework, the scale-space pyramid layers consist of n octaves ci and n intra-
octaves di, for i = {0, 1, . . ., n − 1} and typically n = 4.
The octaves are formed by progressively half-sampling the original image (corresponding to
c0).Each intra-octave di is located in-between layers ci and ci+1 (as illustrated in Figure 5). The first intra-
octave do is obtained by down sampling the original image c0 by a factor of 1.5, while the rest of the intra-
octave layers are derived by successive half sampling. Therefore, if t denotes scale then t(ci) = 2iand t(di) =
2i *1.5 [5].

Figure 5. Scale-space interest point detection

Scale-space interest point detection: a key point (i.e. saliency maximum) is identified at octave ci by
analyzing the 8 neighboring saliency scores in ci as well as in the corresponding scores-patches in the
immediately-neighboring layers above and below. In all three layers of interest, the local saliency maximum
is sub-pixel refined before a 1D parabola is fitted along the scale-axis to determine the true scale of the key
point. The location of the key point is then also re-interpolated between the patch maxima closest to the
determined scale [13].

5.2. Key Point Description

Given a set of key points (consisting of sub-pixel refined image locations and associated floating-
point scale values), the BRISK descriptor is composed as a binary string by concatenating the results of
simple brightness comparison tests. This idea has been demonstrated in to be very efficient, however here we
employ it in a far more qualitative manner. In BRISK, we identify the characteristic direction of each key
point to allow for orientation-normalized descriptors and hence achieve rotation invariance which is key to
general robustness. Also, we carefully select the brightness comparisons with the focus on maximizing
descriptiveness [13].

6. DESCRIPTOR MATCHING
Matching two BRISK descriptors is a simple computation of their Hamming distance as done in
BRIEF: the number of bits different in the two descriptors is a measure of their dissimilarity. Notice that the
respective operations reduce to a bitwise XOR followed by a bit count, which can both be computed very
efficiently on today‟s architectures [13].

7. EXPERIMENTAL RESULTS AND DISCUSSION

7.1. Experimental result with applying algorithm
The experiment is conducted in Android Operating Systems using OpenCv 310 which used image
processing and image comparisons techniques with BRISK algorithm. The database is created of six different
plants. It consists of 30 images of Paddy plant affected Bacterial Blight Disease, Bacterial Leaf Streak
Disease, 20 images of Paddy plants affected by Foot Rot Disease, 38 images of Early Blight and Late Blight

Crops Diseases Detection and Solution System (Md. Abdul Awal)

214  ISSN: 2252-8776

of Tomato plant, 38 images of Early Blight and Late Blight of Potato plant, 20 images of Anthracnose
Disease of Mango plant and 15 images of Rot Wilt Disease of Jute plant.

Table 1. The sample of different stages of affected crops stored into the database

We have used BRISK algorithm in our research to match the uploaded affected crops images with
our database images. As per BRISK algorithm‟s techniques the system takes 100 key objects from both
uploaded image e.g. class 1 and all database images e.g. class 2 for the specific crop then it is compare key
objects between class 1 and class 2. After finding the matching of the objects it does calculate the percentage
of matched found. It follows the below equation:

If calculated percentage is above 70 then display the matched image from database and provide the
possible disease name and solution about the disease otherwise provides message “Uploaded image is not
matched with the database”.

7.2. Result with developed apps

Menu bar is a user interface with different menu item like Dashboard, Take Photo and Upload Photo
etc.

Figure 6. Menu list

IJECE Vol. 6, No. 3, December 2017 : 209– 217

IJ-ICT ISSN: 2252-8776  215

Using Upload Photo menu farmers can upload affected crops photo to detect diseases and will get
solution.

Figure 7. Upload Image Page

After clicking Upload Photo menu item farmers or visitors can see Photo Upload option. Here
clicking users will select specific crops, user can select photo by tapping/clicking blank image and will click
Upload button.

Figure 8. Upload Image Page with process Dialog

After clicking Upload button a process dialog will show.

Figure 9. Successful detection result page

Crops Diseases Detection and Solution System (Md. Abdul Awal)
216  ISSN: 2252-8776

If uploaded photo matched with stored image then successful detection result page will show with
possible disease name, causes and solutions.

Figure 10. Failed detection result page

If uploaded image does not match with stored image, then it will show 'Could not detect disease'
message.

7.3. Comparison statement

We have compared the Crops Diseases Detection and Solution system with others system and got
the better, efficient performance of the system which stated in the below table 2.

Table 2. Comparison of Crops Diseases Detection and Solution with others systems
Criteria Crops Diseases Plant Leaf Disease Detection of Leaf Android Based App to
Detection and Identification System Diseases and Monitoring Prevent Crop Diseased in
Solution system for Android [9] the Agricultural Various Seasons [11]
Resources using Android
App [10]
Enough information stored in Yes Yes Yes Yes
database about plant diseases
Full part of the diseases Yes No No No
Many kinds of plants Yes Yes Yes No
Easy to use Yes Yes Yes Yes
Access from every where Yes Yes Yes Yes
Accuracy Yes No Yes Yes
Faster to get result Yes Yes No No
Details solution Yes No No No

8. ADVANTAGES
a. It is easy to install
b. No need to register
c. It is user friendly
d. It work‟s in different version of android
e. It is light weight
f. It is faster

9. CONCLUSION
This Automated Crops Diseases Detection and Solution System is tested in different Android phone
and it is working nicely and the system is very useful for farmers especially who are living in rural area and
where the agricultural experts are not available. User can easily use this application. Any people can detect
and get solution any time using this application. The application is free of cost and does not require any
additional device.

IJECE Vol. 6, No. 3, December 2017 : 209– 217

IJ-ICT ISSN: 2252-8776  217

REFERENCES
[1] Ayesha Bhandralia, Resham Arya, S. N. Panda, Sachin Ahuja. “Polyhouse Agricultural Marketing System Using
Big Data Hadoop.” International Journal of Advances in Applied Sciences (IJAAS).2016; 5(2):78-84.
[2] Hamid Miah. “Agriculture Sector Development Strategy: background paper for preparation of 7th Five Year Plan,”
http://www.plancomm.gov.bd/7th-five-year-plan/
[3] Hiteshwari Sabrol1, Satish Kumar. “Recognition of Tomato Late Blight by using DWT and Component Analysis,”
International Journal of Electrical and Computer Engineering (IJECE). 2017; 7(1):194-199.
[4] MS Prasad Babu and B Srinivasa Rao. “Leaves Recognition Using Back Propagation Neural Network-Advice For
Pest and Disease Control On Crops,” IndiaKisan.Net: Expert Advissory System. 2007.
[5] Jayamala K. Patil, Raj Kumar. “Plant Leaf Disease Image Retrieval Using Color Moments. IAES International
Journal of Artificial Intelligence (IJ-AI).2013; 2(1): 36-42.
[6] B. Li, et al. “Hyperspectral identification of rice diseases and pests based on principal component analysis and
probabilistic neural network,” Transactions of the CSAE, , 2009;5: 143–147.
[7] H. Jie, “Application of PCA Method on Pest Information Detection of Electronic Nose,” IEEE international
Conference on Information Acquisition. Weihai, Shandong, China, 2006;1456-1468.
[8] M. M. Seinl, et al. “Authentications of Myanmar National Registration Card,” Indonesian Journal of Electrical
Engineering and Informatics (IJEEI), 2013; 2: 53-58.
[9] Prof.H.M.Deshmukh, Jadhav Sanjivani, Lohar Utkarsha, Bhagat Madhuri, Salunke Shubhangi. “Plant Leaf Disease
Identification System for Android,” International Journal of Advanced Research in Computer and Communication
Engineering. 2016; 5: 396-398.
[10] Sufiyan K Shekh, Aniket Baitule, Milind Narethe, Sangappa Mallad, Waghdarikar, Dr D Y Patil. “Detection of
Leaf Diseases and Monitoring the Agricultural Resources using Android App,” International Journal of Innovative
Research in Computer and Communication Engineering. 2015; 3: 9540-9547.
[11] Mr.Sushant. S.Chavan, Mr.Nitesh. P.Satre, Mr.Rajat. R.Deshmane, Mr.Suhas.B.Katkar. “Android Based App to
Prevent Crop Diseased in Various Seasons,” International Research Journal of Engineering and
Technology(IRJET). 2016; 3:1052-1055.
[12] Priyanka L.Kamble, Anjali C.Pise. “Review on Agricultural Plant Disease Detection by using Image Processing.
International Journal of Latest Trends in Engineering and Technology (IJLTET). 2016; 7: 335-339.
[13] Stefan Leutenegger, Margarita Chli and Roland Y.Siegwart. “BRISK: Binary Robust Invariant Scalable Keypoints,”
IEEE International Conference on Computer Vision, ICCV 2011, Barcelona, Spain, .2011;2548-2555.

Crops Diseases Detection and Solution System (Md. Abdul Awal)