Volume 3, Issue 9, September 2013

ISSN: 2277 128X

International Journal of Advanced Research in

Computer Science and Software Engineering
Research Paper
Available online at: www.ijarcsse.com

White Spot Syndrome Virus Detection in Shrimp Images

using Image Segmentation Techniques
M.Muni Sankar*
Assistant Professor,
Dept of ECE, SIET,
Puttur,A.P., India

CH.Nageswar Rao, G.Sailaja

Assistant Professor,
Dept of ECE,SIST,
Puttur,A.P., India

Bhuvaneswary.N
Assistant Professor,
Dept of ECE, APIC,
Poonamallee,Chennai,T.N, India

P.Gunasekhar
Assistant Professor,
Dept of ECE ,R.M.K.E.C
kavaraipettai,Chennai,T.N, India
Abstract- One of the most serious problems confronted by the shrimp farming industry is the disease caused by white
spot syndrome virus (WSSV). This paper describes the detection of white spot syndrome virus in shrimp (Penaeid
prawn) by using K-Means clustering technique in image processing. There are so many methods for finding out wssv
in aqua culture methods in that Highly sensitive capacitive biosensor is one , that uses wssv shrimp pond water and
mixes glutathione-S-transferase tag for white spot binding protein (GST-WBP) was immobilized on a gold electrode
through a self-assembled monolayer. Binding between WSSV and the immobilized GST-WBP was directly detected by
a capacitance measurement. Under optimum conditions the capacitive bio sensor gives the detections in the shrimp
species. This process is time consuming. By using image segmentation techniques we can get accurate results and its
cost is also less and is very fast compared to other techniques. The proposed method consists of capturing image
sensing technique followed by image acquisition, histogram and K-Means clustering methods.
KeyWords: white spot syndrome virus, Image Analysis Algorithm, Camera Calibrations, Segmentation, K-means
clustering
I.
Introduction
White spot syndrome virus (WSSV), the causative virus of the disease, is found in most shrimp farming areas of the
world, where it causes large economic losses to the shrimp farming industry. The potentially fatal virus has been found to
be a threat not only to all shrimp species, but also to other marine and freshwater crustaceans, such as crab and crayfish.
To date, no effective prophylactic treatment measures are available for viral infections in shrimp and other crustaceans.
Due to current aquaculture practices and the broad host range of WSSV, intervention strategies including vaccination
against this virus would be pivotal to save and protect shrimp farming. Several achievements have been attained in the
search of novel vaccines for WSSV. DNA vaccination, recombinant vaccines, oral vaccination techniques and gene
therapy are some of the thrust areas of focus for scientists and researchers. This review article highlights the recent trends
in the detection of WSSV in shrimp either as histogram method or K-means clustering strategies. Gross observations in
shrimp can be easily made at the farm or pond side using little, if any, equipment. Although, in most cases, such
observations are insufficient for prawns species detectors and coastal management. Accurate and detailed gross
observations are need. For hatchery production of penaeid prawn seed a steady supply of spawners is essential for
effectively planning of the operations, it can done by a machine vision system for monitoring prawns in aquaculture
ponds, developed as a tool to study feed consumption and prawn size distributions. Cameras and lighting system operate
in the near infrared to obtain controlled illumination even in the presence of sunlight. Image analysis algorithms for
segmenting prawns are outlined by using refractive index boundaries in underwater imaging and their relevance for
camera calibrations.

Fig1: Penaeid prawn showing all the parts of the body

2013, IJARCSSE All Rights Reserved

Page | 107

Sankar et al., International Journal of Advanced Research in Computer Science and Software Engineering 3(9),
September - 2013, pp. 107-112
Marine aquaculture has become one of the fastest growing industries worldwide, with annual growth rates of
close to 15% for crustaceans during the 2000-2008 period [1]. Over the last decade, some areas of aquaculture, in
particular salmon and tuna farming have benefited from the introduction of farm automation and monitoring equipment,
including video-based sensors ranging from remote feeding and environmental monitoring systems. To date many of
these advances for fish farming have not been implemented in prawn species detection aquaculture. Unlike fish, prawns
spend time on the bottom of ponds foraging and feeding. Husbandry techniques encourage algae blooms and other
sources of turbidity in ponds. The resulting limited visibility combined with the bottom dwelling behavior present
unique challenges, preventing the video techniques commonly used in fish farming from being successful for monitoring
prawn feeding and sizes.
II.
Camera Calibrations
Camera calibration is possibly one of the most classic and fundamental problem in computer vision which has been
studied extensively for decades. It is fundamental because not only every newly produced camera must run calibration to
correct its radial distortion and intrinsic parameters, but also it is the first step towards many important applications in
vision, such as reconstructing 3D structures from multiple images (structure from motion, photometric stereo, structured
lights etc). Camera self-calibration [ 8 ] avoids the use of known calibration pattern and aims at calibrating a camera by
finding intrinsic parameters that are consistent with the geometry[12] of a given set of images. It is understood that
sufficient point correspondences among three images are sufficient to recover both intrinsic and extrinsic parameters.
Algorithms for calibrating a pinhole camera can be primarily classified into two categories; those that require objects
with known 3D geometry, and those that use self-calibration[4,8], including the use of planar calibration patterns. Both
self-calibrations rely on point correspondences across images, it is important for these approaches to extract accurate
feature point locations.
2.1 Camera Lens Calibration
The primarily goal of finding the quantities internal to the camera that affect the imaging process it includes
Position of image center in the image it is typical not at (width/2, height/2) of the image
Focal length
Different scaling factors for row pixels and column pixels
Skew factor
Lens distortion(pin-cushion effect)
Scaling of rows and columns can be based on the camera pixels. These camera pixels are not necessarily square, and
output of the camera may be analog (NTSC). The final image may be obtained by digitizing card i.e., in the form A/D
converter samples NTSC Signal. Below fig shows the complete detail of converting the Captured image into the display
form from into the monitor

Fig2: A/D converter samples NTSC signal.

2.2 Hyper Spectral Imaging
Hyper spectral imaging is an emerging platform technology that integrates spatial information, as regular imaging
systems, and spectral information for each pixel in the image. Compared to conventional RGB imaging, NIR
spectroscopy and multispectral imaging, hyper spectral imaging has many advantages, like containing spatial, spectral
and multi-constituent information and sensitivity to minor components[11]. The combined nature of imaging and
spectroscopy in hyper spectral imaging enables this system to provide images in a three- dimensional (3D) form called
hypercube which can be analyzed to ascertain minor and/or subtle physical and chemical characteristics of a sample as
well as their spatial distributions.
2.3 Image sensing technique system
A laboratory visible and near infrared (VIS_NIR) hyper spectral imaging system was assembled to acquire hyper
spectral images for prawns[7]. As Shown in fig1, the hyper spectral imaging system consists of a imaging spectrograph a
high performance CCD camera an illumination unit containing two 150 W quartz tungsten halogen lamps, a table used
2013, IJARCSSE All Rights Reserved

Page | 108

Sankar et al., International Journal of Advanced Research in Computer Science and Software Engineering 3(9),
September - 2013, pp. 107-112
for samples removing and a computer running the Spectral Cube data acquisition software which controls the motor
speed, exposure time, binning mode, wavelength range and image acquisition. The camera spectral range was from
380nm to 1030nm divided in 512 bands. The camera has
672 X 512 (spatial X spectral) pixels with a spectral
resolution of 2.8 nm.

Fig3: Capturing Image sensing Technique by using Cameras

III.
Prawn Image Segmentation
3.1 Image Acquisition
Glass dish is filled with prawns and was placed on the table in that the prawn is moving continuously and it be
captured using 0.06 s exposure time to build a hyper spectral image with dimensions (x, y ,), where x and y are the
spatial dimensions (number of rows and columns in pixels) and is the number of wavebands. Therefore, the images
were acquired with 672 pixels in x-direction, n-pixels in y-direction (based on the length of the sample) and 512
wavelengths in -direction with 1.23nm between contiguous bands. 100X100 pixels were randomly selected from prawn
image as a region of interest (ROI) and also treated as one sample. These samples were used to extract the spectral
features and Structure features.

Fig 4: Image Acquisition System

The three categories of image-acquisition devices used in image segmentation are (I) Document Scanners (II) Chargecoupled device (CCD) cameras, and (III) Laser based detectors[14]. Document scanners as configured for densitometry
are for measurements on images with one of the colored materials, CBB or silver. They operate in visible array CCD
detectors used with the better densitometers can distinguish adjacent features that are separated by 50 m or greater
(spatial resolution), which is more than adequate for most image segmentation.
3.2 Edge Detection Technique
Edge detection is one of the most commonly used operations in image analysis, and there are probably more
algorithms in the literature for enhancing and detecting edges than any other single subject. The reason for this is that
edges form the outline of an object. An edge is the boundary between an object and the background, and indicates the
boundary between overlapping objects. This means that if the edges in an image can be identified accurately, all of the
objects can be located and basic properties such as area, perimeter, and shape can be measured. Since computer vision
involves the identification and classification of objects in an image, edge detections is an essential tool. In this paper, we
have compared several techniques for edge detection in image processing. We consider various well-known measuring
metrics used in image processing applied to standard images in this comparison. Finally, an inverse transformation is
applied to get the enhanced spatial domain image. Edges of this enhanced image can then be easily found with any
spatial domain technique. Edge detection operators based on max and min operations are available in references [15,9] .
In references [9] the entropy of a fuzzy set defined by an adaptive membership function, over a neighborhood of a pixel
(x,y) is used as a measure of edginess at (x,y). The use of an adaptive membership function makes the detection
algorithm robust. The framework of the algorithm is quite general and works with any measure of ambiguity (fuzziness).
2013, IJARCSSE All Rights Reserved

Page | 109

Sankar et al., International Journal of Advanced Research in Computer Science and Software Engineering 3(9),
September - 2013, pp. 107-112
3.3 K-Means Clustering
Clustering can be considered the most important unsupervised learning problem; so, as every other problem of this
kind, it deals with finding a structure in a collection of unlabeled data.
Clustering is defined as the process of organizing objects into groups whose members are similar in some way. A
Cluster is therefore a collection of objects which are similar between them and are dissimilar to the objects belonging
to other clusters.
The k-means algorithms are an iterative technique that is used to partition an image into k-cluster. In statistics and
machine learning, k-means clustering is a method of cluster analysis which can to portions n observation into k cluster
with the nearest mean[20-5]. The basic algorithms is given below
- Pick k cluster centers either randomly or based on some heuristic.
- Assign each pixel in the image to the cluster that minimum the distance between the pixels cluster centre.
- Re-compute the cluster centres by averaging all of the pixels in the cluster.
Repeat last two steps until convergences are attained. The most common algorithm uses an iterative refinement
technique[6]; due to this ambiguity it is often called the k-means algorithms.
IV. Experimental Results
We taken input image as single prawn image which is captured by using camera and is shown in the fig3, then the
captured image is used to specified the prawn species detectors. This process is also done by taken another two different
types of prawns. The difference in the structure is identified and differentiated by using the following feature extractions.
4.1 Calibration from a Multiple Prawn image for lens distortion
The details of the lens distortion for the prawn image is displayed in MatLab by using orthogonal WDRC which is
shown in the below figure. It can be used to identify the distortion of the original images by comparing with the other
two different prawns. It can be done by using neighborhood technique in K-means clustering algorithm.

Fig4: Original image (Shrimp having white Spot Syndrome Virus)

Fig5: Orthogonal WDRC

4.2 Calibration from a Multiple Prawn Image for noise distortion using wavelet:
The noise distortion of the prawn image is shown in the below figure. The threshold value between the original image
and Histogram adjusted image is taken as o.8
2013, IJARCSSE All Rights Reserved

Page | 110

Sankar et al., International Journal of Advanced Research in Computer Science and Software Engineering 3(9),
September - 2013, pp. 107-112

Fig6: Histogram details by using wavelet

4.3 Image segmentation by using edge detection to indentify the structure extraction:
The below figure gives the complete details of edge detection system which is implemented in MatLab. The given
input Prawn image is compared with other two different Prawns is resulted as three frames i.e., frame1, frame2 and frame
3. Then by using edge detection algorithm implementing in Matlab we can find the nearest feature and structure based on
the movements of prawn to identify whether it is stationary or not.

Frame1

Frame 2

Frame3

4.4 Image segmentation by using K- Means clustering to indentify the diseases distortions:

Fig7: Truly segmented image using K-Means

Fig8: K- Means clustering for detecting any diseases that occurred.

2013, IJARCSSE All Rights Reserved

Page | 111

Sankar et al., International Journal of Advanced Research in Computer Science and Software Engineering 3(9),
September - 2013, pp. 107-112

Fig9: Objects in cluster1, cluster2, cluster3

V.
Discussion and Conclusions
The underwater photogrammetric models for extracting quantitative spatial information of underwater objects using
CCD stereo images have been researched. The integration of multiple sensors for objects measurements will be
conducted. Digital image classification and pattern recognition for specific objects, e.g., prawn species, will be carried
out. In this paper, we are presenting only the detection of the prawn by comparing with another type of prawns by using
image segmentation and clustering algorithms. The presented prototype image sensing system for prawn to addresses
many of the challenges of operating underwater environment. The imaging sensing system is used to capture the prawn
image in the water for our input. The resulted captured prawn image is used for detecting prawn species in the form of
representing the structure of the prawns. We are presenting only identification of prawns by comparing with other
prawns. The image segmentation methods for both features and structured are presented in the results. Our main
contributions are the choosing of proper moments set which gives good feature and structure can be implemented by
using k-means clustering algorithms. This concept is implemented in future for identifying the new prawns or detecting
the wssv prawn species by using the svd and dwt techniques.
References
[ 1 ] UN Food and Aquaculture Organization, The state of world fisheries and aquaculture 2010, 2010.
[ 2 ] INDUCED MATURATION OF PENAEID PRAWNSA REVIEW, M.S. Muthu, a. Laxminarayana.182
[ 3 ] CoNTE, F. S. 1978. Penaeid shrimp culture currentstatus and direction of research. In : Kauai, P. N.C.
J.Sindermann. (Ed.) Drugs and Food from the sea-Myth or reality.
[ 4 ] Boyle, Roger, Vaclav Hlavac, and Milan Sonka (1999) Image Processing, Analysis, and Machine Vision Second
Edition. PWS Publishing
[ 5 ] Bhagwati charanpatel, Dr. G.R.Sinha, an adaptive k-means clustering algorithms for breast image segmentation,
international journal of computer applications(0975-8887), vol 10-n 4, nov-2010
[ 6 ] P. V. G. D. Prasad Reddy, K. Srinivas Rao and S.Yarramalle, Unsupervised Image Segmentation Method based
on Finite Generalized Gaussian Distribution with EM and K-Means Algorithm, Proceedings of International
Journal of Computer Science and Network Security, vol.7, no. 4, pp. 317- 321, April 2007.
[ 7 ] W. Y. Ma and B. S. Manjunath, Edge flow: a framework for boundary detection and image segmentation,
Proceedings of IEEE Conference on Computer Vision and Pattern Recognition, pp. 744- 749, 1997.
[ 8 ] S. J. Maybank and O. D. Faugeras. A theory of self-calibration of a moving camera. The International Journal of
Computer Vision, 8(2):123152, Aug. 1992.
[ 9 ] y.h.Pao, Adaptive Pattern Recognition and Neural Netwroks. Addistion-Wesely, New ,York (1989)
[ 10 ] Norbert Rohrl, Jose Iglesias-Rozas and Galia Xuan, Xiao, Liao,Qingmin, "Statistical structure Weidl, "Computer
Assisted Classification of Brain analysis in MRI brain tumor Segmentation Tumors", Springer Berlin Heidelberg,
pp. 55-60,International conference on image and graphics, 2008. vol.22, pp 421-426, 2007.
[ 11 ] Maxwell, T., 2005. Object-oriented classification: Classification of pan-sharpening quickbird imagery and a fuzzy
approach to improving image segmentation efficiency. MScE Thesis, Department of Geodesy andGeomatics
Engineering Technical Report No. 233 University of New Brunswick, Fredericton, Canada, pp. 157.
[ 12 ] O. Faugeras and G. Toscani. The calibration problem for stereo. In CVPR, 1986.
[ 13 ] F.D., K. Gonzales and N. Deatras. 1981. Survival, growth and production of Penaeus monodon Fab. At different
stockingdens ities in the earthen ponds with flow through system and supplemental feeding. Fish. Res. J. Phils.
6:1-9.
[ 14 ] R. Ramani, Dr. S. Suthanthirvantha, S.Valarmathy, A survey of current segmentation techniques for detection of
Breast Cancer, Image segmentation, October 2012.
[ 15 ] K. Pal and R.A. King, On edge detection of X-ray images using fuzzy set, IEEE Trans, pattern Analysis Mach.
Intell. PAMI-5

2013, IJARCSSE All Rights Reserved

Page | 112

Hitesh Chakravorty, Rituraj Paul & Prodipto Das

Image Processing Technique To Detect Fish Disease

Hitesh Chakravorty

hitesh_chtvy@yahoo.co.in

Research Scholar
Computer Science Department
Assam University,
Silchar, 788 011,Assam,India

Rituraj Paul

riturajpl@gmail.com

Research Scholar
Computer Science Department
Assam University,
Silchar, 788 011,Assam,India

Prodipto Das

prodiptodas@gmail.com

Assistant Professor
Computer Science Department
Assam University,
Silchar, 788 011,Assam,India

Abstract
Disease is one of the major cause for fish mortality. The identification of diseased fishes are at
early stage to prevent and spreading diseases. Manually detecting fish diseases are not error
free. The image of the diseased fish recognise by using PCA method. In this work diseased area
segmentation of fish image based on colour features with K-means clustering. HSV images and
Morphological operation open for better accuracy to diseased area detection and measurement.
Taken four Epizootic Ulcerative Syndrome (EUS) diseased fish images as a case study to
evaluated the proposed approach. The experimental results clear indication of the effectiveness
of proposed approach to improve the diseased identification with greater precision as well as
correctly compute diseased area. The simulation results of this approach is encouraging.
Keywords: PCA, K-Means, HSV, Morphological Operation, Diseased Fish Images, Image
Processing.

1. INTRODUCTION
Fish constitutes a major component of diet for the people of North- East India particularly in
Assam. Fishes form an important element in the village economy. Fishing and fishery is one of
the main source of employment and income for the people residing in the village. Large-scale
mortality occurs among the fresh water fishes has been increased since 1988 due to dreadful
disease Epizootic Ulcerative Syndrome (EUS), symptom tiny red spot on the body surface
initially. This spot later develop ulcer. After few days losing its scales and muscles exposes body.
Infected fish dies within a short period.[1] [2][3].
Fish disease is a serious problem due to its ability to spread rapidly through water to neighboring
aqua-farms. Therefore, rapid and accurate diagnosis is required to control such diseases.
Traditionally, fish diseases have been diagnosed by using the accumulated experience of fisher
man or fishery departmental expert. However, the accuracy of such final diagnosis ultimately
depends on individual skill and experience and the time spent studying each disease. In order to
overcome this limitation, digital image processing technique to detect and classify fish diseases
from digital images. [4][5] [6].

International Journal of Computer Science & Security (IJCSS), Volume (9) : Issue (2) : 2015

121

Hitesh Chakravorty, Rituraj Paul & Prodipto Das

2. MATERIALS AND METHODS

Fish effected with Epizootic Ulcerative Syndrome (EUS) were collected from the different part of
the Barak Valley, Assam and identified by human expert. Pictures(Fig1) of diseased fish were
taken by the SLR camera. Images were 200 by 200 pixels so that an engineering compromise
can be obtained between processing time of algorithms and clarity retention of input images.

(a)

(b)

(c)

(d)

FIGURE 1: (a)Clarias batrachus, (b)Puntius chola, (c)Labeo bata and (d)Labeo gonius are infected with
EUS disease.

2.1 Principal Component Analysis (PCA)

Principal component analysis (PCA) is a statistical procedure that uses an orthogonal
transformation. The PCA approach is used to reduce the dimension of the data by means of data
compression basics and reveals the most effective low dimensional structure of image patterns.
This reduction in dimensions removes information that is not useful and precisely decomposes
the fish structure which involves transformation of number of possible correlated variables into a
smaller number of orthogonal (uncorrelated) components known as Principal Components. Each
fish image may be represented as a weighted sum (feature vector) of the eigen fish, which are
stored in a 1D array. The test image can be constructed using these weighted sums of eigen fish.
When a test image is given, the weights are computed by projecting the image upon eigen fish
vectors. The distance between the weighted vectors of the test image and that of the database
images are then compared. Thus one can reconstruct original image with the help of eigen fish so
that it matches the desired image. [7] [8] [9] [10].
2.2 K-Means Clustering
Cluster analysis, also called segmentation analysis or taxonomy analysis, creates groups, or
clusters, of data. Clusters are formed in such a way that objects in the same cluster are very
similar and objects in different clusters are very distinct. Measures of similarity depend on the
application.
Clustering can be considered the most important unsupervised learning problem; so, as every
other problem of this kind, it deals with finding a structure in a collection of unlabeled data.
Clustering is defined as the process of organizing objects into groups whose members are similar
in some way. A Cluster is therefore a collection of objects which are similar between them and
are dissimilar to the objects belonging to other clusters.
K-Means Clustering is a partitioning method. The function k-means partitions data into k mutually
exclusive clusters, and returns the index of the cluster to which it has assigned each observation.
Unlike hierarchical clustering, k-means clustering operates on actual observations (rather than
the larger set of dissimilarity measures), and creates a single level of clusters. The distinctions
mean that k-means clustering is often more suitable than hierarchical clustering for large amounts
of data. [11] [12] [13] [14] [15].
2.3.1 Feature Extraction In HSV
The HSV space component to reduce computation and improve efficiency. Unequal interval
quantization according the diseased fish color perception has been applied on H, S, and V
components. In accordance with the different colors and subjective color perception

International Journal of Computer Science & Security (IJCSS), Volume (9) : Issue (2) : 2015

122

Hitesh Chakravorty, Rituraj Paul & Prodipto Das

quantification, quantified hue (H), saturation (S) and intensity (V) are obtained. Hue ranges from 0
to 360 degrees, with variation beginning with red, going through yellow, green, cyan, blue and
magenta and back to red [16] [17] [18].
2.3.2 Morphological Operations
Morphological image processing is a collection of nonlinear operations related to the shape or
morphology of features in an image. The morphological operations rely only on the relative
ordering of pixel values, not on their numerical values, and therefore are especially suited to the
processing of binary images [19].

3. EXPERIMENTAL FINDING
3.1 Algorithm for PCA
Let the training set of images be 1, 2 M the average fish of the set is defined by
= 1/

 i
Each fish differs from the average by vector

=

Where i=1.M
The co- variance matrix is formed by
C = . 

Where the matrix A is given by

A = [1,2,.. M]
This set of large vectors is then subject to principal component analysis, which seeks a set of M
orthonormal vectors. To obtain a weight vector W of contributions of individual eigen-fishes to a
fish image, the fish image is transformed into its eigen-fish components projected onto the fish
space by a simple operation.
t

W k=u k
For k=1... M', where M' M is the number of eigen- fishes used for the recognition. The weights
form vector W = [ w1,w2,.wm] that describes the contribution of each Eigen- fish in
representing the fish image, treating the eigen-fishes as a basis set for fish images. The simplest
method for determining which fish provides the best description of an unknown input fish image is
to find the image k that minimizes the Euclidean distance k.
2

k=||(-k)||

Where W k is a weight vector describing the kth fish from the training set.
The MATLAB was used to implement the algorithm(Figure 2) &( Figure 3).

International Journal of Computer Science & Security (IJCSS), Volume (9) : Issue (2) : 2015

123

Hitesh Chakravorty, Rituraj Paul & Prodipto Das

FLOW CHAT FOR EXPERIMENTATION PCA

FIGURE 2: Complete Process of PCA.

International Journal of Computer Science & Security (IJCSS), Volume (9) : Issue (2) : 2015

124

Hitesh Chakravorty, Rituraj Paul & Prodipto Das

FIGURE 3: Fish Recognition using PCA.

3.2 K-means Clustering Algorithm

The basic aim of the proposed approach is to segment colors automatically using the K-means
clustering technique and L*a*b* color space. The introduced framework of defect segmentation
operates in six steps as follows (Figure 4).
Step 1. Read the input image of diseased fish.
Step 2. Transform Image from RGB to L*a*b* Color Space. It consists of a luminosity layer in 'L*'
channel and two chromaticity layer in 'a*' and 'b* channels. Using L*a*b* color space is
computationally efficient because all of the color information is present in the 'a*' and 'b*' layers
only.
Step 3. Classify Colors using K-Means Clustering in 'a*b*' Space. To measure the difference
between two colors, Euclidean distance metric is used.
Step 4. Label Each Pixel in the Image from the Results of K-Means. For every pixel in our input,
Kmeans computes an index corresponding to a cluster. Every pixel of the image will be labeled
with its cluster index.
Step 5. Generate Images that Segment the Input Image by Color. To separate the pixels in image
by color using pixel labels, which will result different images based on the number of clusters.
Programmatically determine the index of each cluster containing the diseased part of the fish
because K-means does not return the same cluster index value every time. But can do this using
the center value of clusters, which contains the mean value of 'a*' and 'b*' for each cluster (Figure
5).
Gray-scale image of Clarias batrachus , Puntius chola , Labeo bata ,Labeo gonius (Figure 6).

International Journal of Computer Science & Security (IJCSS), Volume (9) : Issue (2) : 2015

125

Hitesh Chakravorty, Rituraj Paul & Prodipto Das

FLOW CHAT FOR EXPERIMENTATION K-MEAN

FIGURE 4: Complete Process of K-means Clustering.

International Journal of Computer Science & Security (IJCSS), Volume (9) : Issue (2) : 2015

126

Hitesh Chakravorty, Rituraj Paul & Prodipto Das

(a)

(b)

(c)

(d)

(e)

(f)

FIGURE 5: K-Means clustering (a) Image of Clarias batrachus , Puntius chola , Labeo bata ,Labeo
gonius are infected with EUS disease with five clusters(b) Index Cluster (c) Cluster1, (d)Cluster2, (e)
Cluster3, (f) Blue nuclie.

FIGURE 6: Gray-scale image of Clarias batrachus , Puntius chola , Labeo bata ,Labeo gonius.

3.3 HSV and Morphological Open

Morphological operations open can also be applied to improved Hue images such that their light
transfer functions are unknown and therefore their absolute pixel values are of no or minor
interest. Morphological techniques probe an image with a small shape or template called a
structuring element. The structuring element is positioned at all possible locations in the image
and it is compared with the corresponding neighbourhood of pixels(Fig: 7) &(Fig: 8).

International Journal of Computer Science & Security (IJCSS), Volume (9) : Issue (2) : 2015

127

Hitesh Chakravorty, Rituraj Paul & Prodipto Das

FLOW CHAT FOR EXPERIMENTATION HSV AND PIXELS VALUE

FIGURE 7: Complete Process of HSV and Pixels Value.

International Journal of Computer Science & Security (IJCSS), Volume (9) : Issue (2) : 2015

128

Hitesh Chakravorty, Rituraj Paul & Prodipto Das

(a)

(b)

(c)

(d)

(e)

(f)

FIGURE 8: (a) Image of Clarias batrachus , Puntius chola , Labeo bata ,Labeo gonius are infected with
EUS disease (b) HSV image, (c) Hue image, (d)Improved Hue image, (e) Morphological Open image
and (f) Measured Pixels.

4. RESULTS AND DISCUSSION

Despite the importance of the subject of identifying fish diseases using image processing, the
advances achieved seem to be a little. There is no ideal method would be able to identify any
disease in any kind of fish. The present approach step forward to achieving fish disease
identification and prevention.
The images of Epizootic Ulcerative Syndrome (EUS) diseased fish such as Clarias batrachus,
Puntius chola, Labeo bata ,Labeo gonius features are extracted and processed by PCA to form
the feature vector after that classify them according to the Euclidian distance. Experimental result
indicates that the algorithm is workable with an accuracy greater than 90 percent.PCA used
successfully in fish pathogen detection and face recognition.
Figure 5 shows the defect segmentation result of an diseased fish of Clarias batrachus , Puntius
chola , Labeo bata ,Labeo gonius using K-means clustering technique. The segmented input
image into five clusters in Figure 5 and it is clear that Cluster1,2 and 3 correctly segment the
diseased portion. In this experiment input images are partitioned into Cluster 1,2 and 3 yields
good segmentation as per requirement. K-Mean clustering algorithms technique used
successfully in shrimp having white Spot Syndrome Virus, infected fruit part detection, detection
of the plant diseases, skin colour detection and classification of leaf diseases.

International Journal of Computer Science & Security (IJCSS), Volume (9) : Issue (2) : 2015

129

Hitesh Chakravorty, Rituraj Paul & Prodipto Das

Figure 8 HSV diseased fish images show the diseased area clearly. Hue indicates the dominant
color of fish diseased area. The morphological operation open on a improved Hue image creates
a new image in which the pixel has a non-zero value and successfully indicate the diseased area
as well pixels value. The experimental results suggest that the introduced method for diseased
segmentation accurately segment the diseased portion and measured pixels value of Clarias
batrachus , Puntius chola , Labeo bata ,Labeo gonius. HSV used successfully in skin colour
detection and texture feature extraction. The morphological operation used successfully in fish
pathogen detection and brain tumors extraction from MRI images.

5. CONCLUSION
The proposed approach to investigate diseased fish through simulation to achieving the
information to identified diseased fish of the fishery in the village area. It also showing automatic
image of diseased fish from the fish database, helps to take curative and preventive measured to
the spread of disease between fish farms rather than general diagnosis and rely on the
performance of a human expert, leads improve management strategies in the village fish farming
industry.
The image processing technique to detect fish disease is a complex problem demand more
sophisticated approaches. For further study need to improve the performance, robustness and
accuracy using techniques like pattern recognition of diseased area of the infected fish using
neural networks, genetic algorithms, support vector machines and fuzzy logic.

6. REFERENCES
[1]

D. Kar, S.C.Dey and A. Roy, 2000. Epizootic Ulcerative Syndrome In Fish at Barak Valley,
Assam, India, Section-8 Sustainable Water Resource Management, Policies And
Protocols.

[2]

D.Kar, A.V.Nagarathna, T.V.Ramachandra and S.C.Dey 2000.Fish Diversity and

Conservation
Aspects in an Aquatic Ecosystem in Northeastern India, Zoos Print
Journal 21(7):2308-2315.

[3]

http://assamagribusiness.nic.in/2ndoct/waterbodiesofAsam.pdf.nic.in.

[4]

J.G.A.Barbedo ,2013: Digital image processing techniques for detecting, quantifying and
classifying plant diseases. Springer Plus 2013 2:660.

[5]

S.W.Lee,J.S.Park,2011.,"MicroscopicImageRecognitionbasedFishDiseaseDiagnosisSyste
m,Journ alofConvergenceInformationTechnology(JCIT) Volume6, Number10, October
2011
doi:10.4156/jcit.vol6.issue10.45.

[6]

J.S.Park, M.J.Oh, S.Han, 2007. "Fish Disease Diagnosis System Based on Image
Processing of Pathogens' Microscopic Images," fbit, pp.878-883, 2007 Frontiers in the
Convergence of Bioscience and Information Technologies, 2007.

[7]

A.Banerjee, 2012. Impact of Principal Component Analysis in the Application of Image

Processing, International Journal of Advanced Research in Computer Science and
Software Engineering, Volume 2, Issue 1, January 2012 ISSN: 2277 128X.

[8]

H. Kadiya ,2012Comparitive Study on Face Recognition Using HGPP, PCA, LDA,ICA and
SVM, Global Journal of Computer Science and Technology Graphics & Vision,Volume 12
Issue 15 Version 1.0 Year 2012,Type: Double Blind Peer Reviewed International Research
Journal,Publisher: Global Journals Inc. (USA) ,Online ISSN: 0975-4172 & Print ISSN:
0975-4350.

[9]

K. Singh Sodhi , M. Lal,.2013 Comparative Analysis of PCA-based Face Recognition

System using
different Distance Classifiers, International Journal of Application or

International Journal of Computer Science & Security (IJCSS), Volume (9) : Issue (2) : 2015

130

Hitesh Chakravorty, Rituraj Paul & Prodipto Das

Innovation in Engineering & Management (IJAIEM), Volume 2, Issue 7, July 2013 ISSN
2319 4847.
[10]

P. Dhoke, M.P.Parsai, 2014 A Matlab based Face Recognition using PCA with Back
Propagation Neural network, International Journal of Innovative Research in Computer
and Communication Engineering,(An ISO 3297: 2007 Certified Organization) Vol. 2, Issue
8, August 2014, ISSN(Online): 2320-9801,ISSN (Print): 2320-9798.

[11]

M.Muni Sankar, CH.Nageswar Rao, G.Sailaja , N.Bhuvaneswary, P.Gunasekhar ,. 2013

White
Spot Syndrome Virus Detection in Shrimp Images using Image Segmentation
Techniques, International Journal of Advanced Research in Computer Science and
Software Engineering, Volume 3, Issue 9, September 2013 ISSN: 2277 128X.

[12]

R.S.Dubey, P.Dixit,N.Singh, ,J.P.Gupta,2013 Infected Fruit Part Detection using K-Means

Clustering Segmentation Technique, International Journal of Artificial Intelligence and
Interactive Multimedia, Vol. 2, N 2.

[13]

S.Naikwadi, N.Amoda, 2013 Advances in Image Processing for Detection of Plant

Diseases, International Journal of Application or Innovation in Engineering & Management
(IJAIEM),Volume 2, Issue 11, November 2013 ISSN 2319 - 4847.

[14]

R.Vijayanandh, G.Balakrishnan, 2011 Performance Analysis of Human Skin Region

Detection Techniques with Face Detection Application, International Journal of
Modeling and Optimization, Vol. 1, No. 3, August 2011.

[15]

V.Ashok, D.S.Vinod, 2014 Using K-Means Cluster and Fuzzy C Means for Defect
Segmentation in Fruits, International Journal of Computer Engineering & Technology
(IJCET),ISSN 0976 6367(Print),ISSN 0976 6375(Online), Volume 5, Issue 9,
September (2014), pp. 11-19.

[16]

D.A.Bashish, ,M.Braik and S.Bani-Ahmed, 2011Detection and Classification of Leaf

Diseases using K-means-based Segmentation and Neural-networks-based Classification,
Information Technology Journal 10 (2) : 267 - 275, 2011. ISSN 18125638/DOI:10.3923/itj.2011.267.275 @ 2011 Asian Network for Scientific Information.

[17]

P.M.Patil, and Y.M.Patil, 2012 Robust Skin Colour Detection And Tracking Algorithm
International Journal of Engineering Research & Technology (IJERT) Vol. 1 Issue 8,
October 2012,ISSN : 2278-0181.

[18]

M.M.Lumb, and P.Sethi, 2013 Texture Feature Extraction of RGB, HSV, YIQ and Dithered
Images using GLCM, Wavelet Decomposition Techniques ,International Journal of
Computer Applications (0975 8887) Volume 68 No.11, April 2013 .

[19]

R .C.Patil, and A. S.Bhalchandra, 2012 Brain Tumour Extraction from MRI Images Using
MATLAB, International Journal of Electronics, Communication & Soft Computing Science
and Engineering ISSN: 2277-9477, Volume 2, Issue 1.

International Journal of Computer Science & Security (IJCSS), Volume (9) : Issue (2) : 2015

131