Download Thunderstorm Detection and its Equivalent Knowledge

ISSN No: 2309-4893
International Journal of Advanced Engineering and Global Technology
I
Vol-05, Issue-01, January 2017
Thunderstorm Detection and its Equivalent Knowledge Representation
with Wavelength Based Image Segmentation
Miss Vasanti Y. Gaud
Prof. Mr. S. S. Kulkarni
M.E student
Department of Information Technology
Prof Ram Meghe Institute of technology
& Research,
Badnera-Amravati (INDIA)
Email-vasantigaud31@gmail.com
Associate professor
Department of Information Technology
Prof Ram Meghe Institute of technology
& Research,
Badnera-Amravati (INDIA)
Email-skulkarni@mitra.ac.in
Abstract
Thunderstorm is a vicious, climatic
disturbance that is associated with heavy
rains, lightening, thunders, thick clouds and
gusty surface winds. Thunderstorms take
place when a layer of warm and moist air
rises to a larger extent, and updrafts to the
cooler regions of the atmosphere. The
updraft that contains moisture condenses in
order to form massive cumulonimbus clouds
and eventually leads to the formation of
precipitation. Columns of frozen air then
sink earthward, striking the ground with
strong downdrafts and horizontal winds.
Meanwhile, electrical charges mount up on
cloud particles and causes lightning. This
further heats the air in a fierce manner by
which shock waves are produced, resulting
in thunder. Thunderstorms were predicted
based on the severity of the sounds of the
thunder, statistical test and graphing were
the other parameters used for the prediction
purpose. In this, many of the researchers
proposed various methodologies like STP
model, MOM model, CG model, LM model,
QKP model, DBD model and so on for the
detection. The research work adopted
clustering and wavelet transform in order to
improve the thunderstorm prediction rate.
The study carried on the thunderstorm
prediction using clustering and wavelet
techniques resulting with higher accuracy.
Keywords – Clustering, wavelet transform,
Image processing, Thunderstorm .
I.
Introduction:
A climatic disturbance that is associated
with heavy rains, lightening, thunders, and
thick gusty and clouded surface wind is
known as thunderstorm. Thunderstorms take
place when a layer of warm and moist air
rises to a larger extent, and other gases to the
cooler regions of the atmosphere.
The other gas that contains moisture
condenses in order to form large mass
cumulonimbus clouds and lead eventually to
the formation of precipitation. Columns of
frozen air then sink earthward, striking the
ground with strong downward air and
horizontal winds. Meanwhile, electrical
charges rise up on cloud particles and causes
1492
www.ijaegt.com
ISSN No: 2309-4893
International Journal of Advanced Engineering and Global Technology
I
Vol-05, Issue-01, January 2017
lightning. This further heats the air in a
fierce manner by which shock waves are
produced, resulting in thunder.
Cloud lightning frequently occurs as part of
the thunderstorm phenomena, which on
rigor becomes dangerous to the property,
wildlife and population across the world to a
major extent. One of the most significant
lightning hazards is to the wildfires, as they
can even burn the ground surfaces. The
behaviour of thunderstorms is subjected to
the experience of the forecasters and the
analysis of numerical. Thunderstorms
prediction based on the harshness of the
sounds of the thunder, analytical test and
graphing were the other parameters used for
the prediction purpose.
Thunderstorms are generally formed in any
geographic location. The satellite images
obtained from Indian Meteorological
Department, which predict whether the
cloud images produce thunderstorms or not.
The input image of the cloud is been taken
by the satellite for the experimentation. The
image taken by the satellite is the input
image.
The rapid upward movement of warm
shows the result of thunderstorm, moist air,
this can be represented in 3 stages named as
the developing, maturity, and dissolving
stage. The developing stage is when, the
storm starts strengthening in this the warm,
moist air rises above and gets mixed with
the freeze air making the warm air to get
colder resulting in condensation. In this
stage, the cloud forms larger due to the
instability in the atmosphere and moves to
the next stage.
The maturity stage starts when the storm
reach its peak and is well developed,
including a strong, dense anvil along with
updrafts and downdrafts and in this stage
hail may also prevail When a storm does
this, it means the storm is very strong and
has capability to produce severe weather and
tornadoes and in the last stage called the
dissolving stage. The storm starts fading
away, when the cool downdrafts begin to
intensify, the storm begins to dissolve.
These downdrafts basically push everything
out of the storm.
Thunderstorm consists of the four types
single cell, multi cell cluster, multi cell line
and super cells. Where super cell
thunderstorms are strongest and associated
with severe weather.
II.
Literature review:-
Himadri Chakrabarty, C.A Murthy, S.
Bhattacharya and A. D. Gupta [1] used
Artificial Neural Network model which
predict Squall thunderstorms by using
RAWIND data. Litta A.J, Sumam Mary
Idicula and Naveen Francis C [2] adopted
multilayer perceptron network model to
predict thunderstorms where in, the
prediction was done using the data obtained
from RSRW flight but this is limited to a
particular region rather than the entire world
prediction over Kolkata. Harvey Stern [3]
used a knowledge based system to predict
thunderstorms. Tajbakhsh S, Ghafarian P,
and Sahraian F [4] adopted numerical
weather prediction model in order to survey
thunderstorms. Pinto [5] discussed annual
values of thunderstorm days
and
temperature values in the city of Sao Paulo.
Alan Czarnetzki C [6] discussed about
nocturnal thunderstorms, which produces
heavy rains. David Bright R, Matthew
1493
www.ijaegt.com
ISSN No: 2309-4893
International Journal of Advanced Engineering and Global Technology
I
Vol-05, Issue-01, January 2017
Wandishin S, Ryan Jewell E and Steven
Weiss J [7] used a physically based
parameter for lighning prediction. Ken
Harding [8] discussed the formation of
thunderstorms and its aviation hazards. Bill
Nisley [9] discussed the formation and
anatomy of thunderstorms.
To segment the input image into several
clusters based on similarity measure where
clustering is an efficient technique, one of
the similarity metric used is Euclidean
distance. For segmenting the image k-means
clustering is adopted. Based on various
colour factors segmentation is performed to
image because of colours possess
wavelength values. The segmentation
process applied for a thunderstorm satellite
image, and the resulted clusters generated.
The feature extracted clustered image is
analyzed further by applying wavelet
transformations.
The
Haar
wavelet
transform is adopted for the further analysis
where columns by transforming from data
space to wavelet space in frequency domain
and decomposition is applied to the image in
rows. The main aim is to detect the
thunderstorms as accurate as possible.
III.
Third stage is thunderstorm detection, which
also provides wavelength calculation. After
the calculation of wavelength it detect the
lightning thunderstorm region, as these
wavelength calculations can detect whether
it is thunderstorm or not.
The first stage is to recognize the lightning
thunderstorm image. When the input images
consist of the thunderstorm image and
background image. As these image consist
of some noise section, by cropping these
thunderstorm image noise is been get
removed. As the lightning thunderstorm
image consists of the region of interest and
minimum of background image.
1. Database creation
Start
Input Image
K-Means Clustering
Wavelet Transform
Proposed Methods:
The proposed system consist of the three
stages explained below.
This first stage is to recognize the lighting
area with the help of input image and also
enhance the image for detection of
thunderstorm.
Second stage is the clustering where by
using the k-means clustering algorithm we
add the accuracy in this stage. It will get the
more precise lighting thunderstorm image so
that it can easily identify.
Image Segmentation
End
Fig1:- Database Creation
The database creation stage which consists
of the k-means clustering and wavelet
transformation of the image; it is an efficient
technique to divide the input image into
several cluster. Take the input image of
1494
www.ijaegt.com
ISSN No: 2309-4893
International Journal of Advanced Engineering and Global Technology
I
Vol-05, Issue-01, January 2017
lightning thunderstorm where the k means
clustering is done. Where the thunderstorm
image consist of the noise, which is been
remove by enhancing the thunderstorm
image. While clustering is been done to get
the perfect thunderstorm image after the
deletion of the noise. The wavelet
transformation is done to enhance the image.
Segmentation is based on different values
grouped into the different clusters. Edge are
been get calculated in these segmentation,
where the edge of thunderstorm is
calculated. And the thunderstorm is been
stored into the database creation.
unwanted noise and enhance thunderstorm
image. Image segmentation is done to
calculate the edges of the thunderstorm.
While this evaluation of edges, the
thunderstorm will be detect. Thunderstorm
detection will show the pattern of the
lightning thunderstorm form the detected
thunderstorm image. After the detection of
lightning thunderstorm the knowledge
representation is done. This shows the data
information. It represents the formation of
information.
Result:-
2. Thunder Strom Detection
Start
Input Image
Image Segmentation
Fig 3:-Screen Shot
Wavelength Calculation
Thunder Strom Detection
Knowledge
Representation
Fig 2:- Thunder Strom Detection
The next stage is testing stage where the
thunderstorm image is been tested for the
detection. Whether these input image is
thunderstorm or not. Take the input
lightning thunderstorm image from the
database. Enhance this input image of
lightning thunderstorm, which extract the
Fig.4:-Screen Shot
IV.
Conclusion:
As we have proposed the thunderstorm
detection, where these detection of
1495
www.ijaegt.com
ISSN No: 2309-4893
International Journal of Advanced Engineering and Global Technology
I
Vol-05, Issue-01, January 2017
thunderstorm image is been done. The
thunderstorm image is been enhanced. The
extract region of interest of images which
gives the perfect image of thunderstorm and
get the section of thunderstorm. The
wavelength calculation is been done and get
the segmented image of thunderstorm. This
detection of thunderstorm by wavelength
based image segmentation and knowledge
representation is also been done. The
projected method predicts 75% of the
thunderstorm more accurately than the
previous method. In future work, weather
models can predict the future state of the
atmosphere they must know its present state
very accurately. It is the primary purpose of
a disparate array of weather measurement
instruments and platforms and various other
special-purpose devices.
V.
Reference :
[1] Himadri Chakrabarty, Murthy, C. A., Sonia
Bhattacharya and Ashis Das Gupta,
“Application of Artificial Neural Network
to Predict Squall-Thunderstorms Using
RAWIND Data,” International Journal of
Scientific and Engineering Research, 2013,
pp. 1313-1318.
[2] Litta, A.J., Sumam Mary Idicula and
Naveen Francis C, “Artificial Neural
Network Model for the Prediction of
Thunderstorms over Kolkata”, International
Journal of Computer Applications, 2012,
pp. 50-55.
[3] Harvey Stern, “Using A Knowledge based
System to predict Thunderstorms,” Bureau
of Meteorology, Australia.
[4] Tajbakhsh, S., Ghafarian, P, and Sahraian,
F., “Instability Indices and Forecasting
Thunderstorms: the case of 30 April 2009,”
Natural hazards and Earth System Sciences,
2012, pp. 403-413.
[5] Pinto,
“The
Sensitivity
of
The
Thunderstorm Activity in the city of Sao
paulo to temperature Changes: predicting
the Future Activity for Different
Scenarios,”
International
Lightning
Detection Conference, 2012, pp. 1-4.
[6] Alan czarnetzki, C., “Evaluation of a
Forecast
strategy
for
Nocturnal
Thunderstorms that Produce heavy rain,”
pp. 25-31.
[7] David Bright, R., Matthew Wandishin, S.,
Ryan Jewell, E., and Steven Weiss, j., “A
Physically Based Parameter for Lightning
Prediction and its Calibration in Ensemble
Forecasts,” Confeerence on Meteorological
Applications of Lightning Data, AMS,
2005, pp. 1-11.
[8] Ken Harding, “Thunderstorm Formation
and Aviation Hazards,”NOAA’s National
Weather Service, pp.1-4.
[9] Bill Nisley, “Thunderstorm Anatomy and
Dynamics,”Naval Postgraduate School,
california, 1999, pp. 1-13.
[10] Kanwaljot Singh Sidhu, Baljeet Singh
Khaira, Ishpreet Singh Virk, “Medical
Image Denoising In The Wavelet Domain
Using Haar and DB3 Filtering,”
International
Refereed
Journal
of
Engineering and Science, 2012, pp. 1-8.
[11] Candra Dewi, Mega Satya Ciptaningrun,
Nmuh Arif Rahman, “Denoising Cloud
Interference on Landsat Satellite Image
Using
Discrete
Haar
Wavelet
Transformation,” International Journal of
Computer Science and Information
Security, 2012, pp.27-31.
[12] C. Fraley and A. Raftery, "How Many
Clusters? Which Clustering method?
Answers
via
Model-Based
Cluster
Analysis," The Computer Journal, 1998, pp.
578-588.
1496
www.ijaegt.com