......

..

IJCTA-Volume 8 Issue 3 / MAY-JUNE 2017
S.No
Title/Author Name
Page No
1
Automatic Number and license plate Recognition system
-Praveen Kumar,Chaitanya MSVS,M.Sailesh,G.Sai Siril
Abstract
In this paper, we are proposing a mechanised and automatic number and licence plate recognition system which can extract the characters from the number plate of the vehicle passing through a location where the system has been installed using image processing algorithms. There is no need of additional devices like GPS, Bluetooth etc. to be installed for the proposed system. Only RFID and the usage of camera has been implemented for the additional security purpose. Using RFID detects the identification of the person and using of cameras, takes the pictures of the passing vehicles and forwards the image to the computer for being processed, the software used uses different algorithm like normalisation, segmentation and optical character recognition (OCR).The resultant data is compared with a data in the data base Experimental results show that the proposed automatic number plate recognition system successfully recognizes and detects the number plate of the vehicle whose details has to be verified. This system can be used for security, parking and traffic purposes.
Keywords: LNPR, OCR, RFID, GPS.
302-305
PDF
2
Real Time Stream Processing in a Big data with Hadoop
-Goutham R,Saravana Kumar K.R.Subash
Abstract
Real-Time Stream Processing In A Big Data With Hadoop(RSPBH), The remote sensing virtual international daily generate large extent of real-time records, data are remotely amassed inclusive of to investigate, aggregate, and stock it. huge statistics is extracting the beneficial information in an efficient manner. The Analytical Architecture overcomes these challenges by using 3 main units SDU, SDPU, DMU-CA. In Streaming data unit(SDU) the data is collected and placed in the local file of the host machine. In Streaming Data process unit(SDPU) the main reduction of data takes place by Mapreduce technique present again in Hadoop. In Decision making unit for complex analysis (DMU CA) the data stored after the Mapreduce is placed back in local host machine as we don’t have analysis option in Hadoop.
Keywords: Big data, SDU, SDPU, DMU-CA,RSPBH
306-310
PDF
3
A Novel IOT Approach for Human Presence Detection and Home Monitoring System
-Reenu Varghese,Saju A
Abstract
Due to hardware limitations of Raspberry pi, motion detection is implemented by comparing the last two consecutive images saved locally. In this work, can be identified intruder motion there by securing the area. When a face is identified, the algorithm will check the database. If the face is in the database, it will show whose face is it. It the face that are identified is not in database, it will show as “unknown”. At that particular instant of time, the server notifies to the user and all the images from the point of identification are uploaded to a cloud server and sent via email. Through the website we can observe the room and control operations can be done by raspberry pi. This works aims to describe a security system using low processing raspberry pi using Internet of things.
Key Words –IOT, Raspberry Pi, Webcam
311-315
PDF
4
WikiSum - A Text Summarization Application for Wikipedia Documents
-Renuka Devi,Aditya B,Abhinav Rajput
Abstract
We display a technique for separating sentencesb from an individual report to fill in as an archive outline or a pre-cursor to making a nonexclusive record unique. We apply syntactic investigation of the content that delivers an intelligent frame examination for each sentence. We utilize subjectobjectpredicate (SOP) triples from individual sentences to make a semantic diagram of the first report and the relating human separated rundown. Utilizing the Support Vector Machines learning calculation, we prepare a classifier to distinguish SOP triples from the archive semantic chart that have a place with the rundown. Due to the consistent development of Wikipedia,WikiSum likewise gives an expanding increased the value of information procurement, re-utilize and joining assignments inside organisations. During the advancement of WikiSum we discovered that it is critical to handle Wikipedia refreshes in a need line. As of late refreshed Wikipedia articles ought to have the most elevated need, over mapping-changes and unmodified pages. A general finding is that there is an a lot of chances emerging from the developing Web of Data for administrators.
Index Terms—SDN, OpenDaylight, Relay , LATEX, paper, template.
316-320
PDF
5
Enhancement of Security in VANET by Detection of Malicious Nodes using Trust System
-Diljot kaur,Simarjit Singh Malhi
Abstract
Vehicular ad-hoc network (VANET) is one of the recent and promising technologies to revolutionize the transportation system where vehicles can communicate by exchanging messages via wireless medium. It has received a lot of interest in the last few years. But still there are many issues that need to be resolved for its betterment and use in the transportation system. Though, security is of major concern whenever we talk about communication between entities in a network, therefore it needs to be addressed perfectly. This paper present a security enhancement which when deployed will produce efficient results in the detection and prevention of malicious nodes in the system. With the detection of malicious nodes in the system, we will reduce the risk by selection of the trusted nodes and hence will increase system availability.
321-326
PDF
6
Enhancement of Routing in Urban Scenario using Link State Routing Protocol and Firefly Optimization
-Dhanveer Kaur,Harwant Singh Arri
Abstract
Vehicular ad-hoc network (VANET) is one of the recent and promising technologies to revolutionize the transportation system where vehicles can communicate by exchanging messages via wireless medium. It has received a lot of interest in the last few years. But still there are many challenges that need to be resolved for its efficient use in the transportation system. Though, in case of selection of street at the intersection and relay node selection in intra-street is of major concern whenever we talk about communication between entities in a network, therefore it needs to be addressed perfectly. This paper present a routing enhancement which when deployed will produce efficient results in the intra-street communication and selection of street at intersection. With the selection of optimal relaying nodes, we will improve the ratio of packet delivery, network yield, reduce the end to end delay, and reduce the routing overhead.
327-333
PDF
7
A Contemporary Approach to Component-Based Software Testing
-Saurabh Rawat,Rajesh Kumar
Abstract
Testing is a vital action encouraging productivity, not negligible, in the sphere when Software Development community is enveloped by Component Based Software Development. Testing constitutes more than fifty percent of the expense of Software Development, resulting in increased cost.This paper describes the issues and challenges of component-based systems. This paper also suggests the requirement of a novel approach to testing of component-based software. It also proposes a novel approach to testing of component-based software while considering some important factors like component study, component test case design, component test execution, and component test analysis and component test documentation.
Keywords - Component-based development, Software Component Testing, Issues and Challenges in Component-Based Software Testing
334-337
PDF
8
Optimized FIR Filter Using Distributed Parallel Architecture for Audio Application
-Prathibha P Nair,Tintu Mary John
Abstract
Digital Signal Processing is an important area having a wide range of applications with Digital Filters as the key element. Different architectures can be used for implementing FIR filters, Distributed Arithmetic is one such architecture. This is a multiplier less architecture, since multipliers are the speed limiting elements in a VLSI circuit. Thus it consumes less area than basic FIR filter by replacing multipliers with shift and add operation. Multimedia requires signal processing to be fast with less power consumption. Audio signal processing is an important application in multimedia. In order to perform this filtering function efficiently, we can optimize the delay, power and area, for that, Distributed Parallel Architecture is used. In this paper an FIR filter is designed using the Filter Design and Analysis (FDA) tool in MATLAB and the filter coefficients are calculated in 16-bit fixed point format. The input values of the audio signal is generated to HDL code by programming in MATLAB R2015a. The filter is implemented for Audio Application using basic filter structure, serial Distributed Arithmetic and parallel Distributed Arithmetic Architecture in Xilinx ISE 14.7 by programming in VHDL and the speed, area and power of these filters are compared. 40% increase in speed is obtained in Parallel Distributed Arithmetic FIR filter compared to the serial structure and when compared to the basic Direct form structure, there is 70% speed improvement.
Keywords- Digital Signal Processing, FIR filter, Distributed Arithmetic, Distributed Parallel filter
338-342
PDF
9
Web based Data Acquisition and Control System for Thermal Power Plants
-Soniya Sunny,Kuruvilla John
Abstract
Wireless sensor network technology is emerging as an important field in process monitoring and controlling applications. In thermal power plants, various environmental conditions have to be continuously monitored. This paper presents a web-based data acquisition and control system for thermal power plants. The system is implemented using Raspberry Pi and ZigBee protocol. Different sensor nodes are placed at different locations in the power plant. A coordinator node is designed with Raspberry Pi as the main controller. All sensor nodes are interconnected using ZigBee communication protocol. Raspberry Pi communicates with all sensor nodes in the wireless sensor network and collects the values sensed by these sensor nodes. It will then upload all information to the internet so that the user can monitor data from distant location through web browser. A threshold value is set at sensor nodes, and if the sensed value crosses the threshold limit, a buzzer will beep in order to provide control signal.
343-348
PDF
10
Accuracy of Telugu Language Speech Recognition corresponding to Size of the Speech Corpus Database
-P.Jeethendra,M.Chandrashekar
Abstract
In this paper, efforts were put in finding development of Indian language speech databases in general and Telugu in particular for building large vocabulary speech recognition systems. The speech to text database for the Speech and Hearing Disabled needs a higher order of accuracy and wide database and a high word recognition rate with lowest error. The design and methodology of collection of various speech databases is discussed.
One of the major challenges in the field of Automatic Speech Recognition (ASR) is Modeling of rich transcription, modeling. The speech database should be rich in many dimensions, such as in text, environments, transducer type, number of recording sessions, recording system, the transmission channel, the country of origin, and the mother language. Research in any Language processing and very useful in many speech processing tasks, such as speaker recognition, speech recognition, and accent identification.
It was found that databases for European Language are widely and abundantly available (2) but a little database is available for Indian Languages. The accuracy of the speech recognition depends upon the size and amount of database in that language.
Keywords : Database, Large Vocabulary, Automatic Speech recognition, Accuracy, Indian Languages.
349-360
PDF
11
Classification and Model Evaluation Using Hyperspectral Data of Vegetative Landscape
-V Merin Abraham,Jacob Thomas
Abstract
Species classification is an important task when it comes to the study of specific areas without disrupting the ecosystem. In this paper we chose Indian Pines as the test site to classify it into 16 classes of tree species and compare them with ground truth to obtain the accuracy, using various algorithms with machine learning process. The algorithms used are LDA and SVM. Spectral features of trees are obtained from hyperspectral data. Results show improvement in classification performance. We obtain an accuracy of 82 percent for LDA and 84 percent for SVM. We also try to implement spectral unmixing for the same set of data. Usage of spectral unmixing leads to endmember extraction method thus identifying the constituent components of a pixel which can be of further used by scientist and researchers. Our aim is to indicate which algorithm works best for test site.
Keywords-Species classification, machine learning, SVM, LDA, hyperspectral data, ground truth
361-366
PDF
12
User Authentication Using Keystroke Dynamics based of Selected Features
-Sandeepan Dhoundiyal,Dhruv Shukla,B. Ida Seraphim
Abstract
Keystroke dynamics is the time analysis of a users typing habit. Keystroke dynamics is a biometric method of user authentication. However an authentication technique must strive to balance two properties: speed; accuracy. In order to balance these conflicting properties, a small set of accurate features and a simplistic measurement technique. These are the measures of flight time, dwell time, and latency time, and the technique of confidence interval based statistical analysis.
Keywords—Keystroke dynamics; Authentication; Selected parameters.
367-369
PDF
13
Agricultural Marketing Information System – Architecture, Framework and Model
-Dr. Syed Khizer
Abstract
The process of marketing, finding storage and planting advice is a challenge to farmers. Therefore, there is a need for more reliable, efficient and automated system to help farmers on finding information on storage, marketing and accessing planting advice using present-past agricultural marketing data. Agriculture marketing data plays a very important role in advisory, planning, management and control of agriculture commodities in a country. In this research paper we present a model which focuses on different aspects of developing information system for managing agricultural marketing data. The model presented here can be used to develop Agricultural Marketing Information System (AMIS) and similar systems.
Keywords: Information System, UML Modelling, Advisory System, Agricultural Marketing System, Business Intelligence, Data Management, Data warehousing, Information management.
370-374
PDF
14
Reputation Based Ranking System of URLs
-Sudhanshu Kumar Tetarbe, Saumya Goyal, Mihir P, Dr.Revathi,Venkataraman
Abstract
Security is an issue that has gripped the Internet community for quite some time now. Be it be phishing attacks, Denial of Service attacks or any other type of cyber-attacks, the Internet is in a need of a system which can help users gauge the threat level of a website. This is where our reputation based domain ranking system comes into play. An initial list of malicious websites is generated from three different trusted sources. Each of these sources use a different unique algorithm to judge a website to be malicious. The URLs are fed to a program which shall calculate the 10 parameters for each of these URLs by which we calculate the weight for each parameter. Once the weights are calculated, reputation point is calculated based on a mathematical formula. Based on the reputation points calculated, the initial ranking is done for all the websites in the data set. Now, a web interface is hosted on Internet, where the user may enter the URL they want to check the rank of, and our system returns the ranking of the URL. Also, our database is updated dynamically whenever a new URL is entered in the web interface.
Keywords: Security, Cyber-Attacks, Reputation point, Domain Rankin
375-381
PDF
15
Face Detection using NN & Gabor Wavelet Transform
-Krati Sharma,Sapana Singh
Abstract
This paper is based on classifications of the features of a face detected using Gabor filter feature extraction techniques in image processing. paper focuses on feature extraction in an image using Gabor filter and the extracted image feature vector is then given as an input to the neural network. The neural network is trained with the input data. The Gabor wavelet concentrates on the important components of the face including eye, mouth, nose, cheeks. The main requirement of this technique is the threshold, which gives privileged sensitivity. These feature vectors are given into the feed forward neural network to train the network. Using the feed forward neural network as a classifier, the recognized and unrecognized faces are classified. The effectiveness of the proposed method is demonstrated by the experimental results on testing large number of images and comparisons with state of the art method. effectiveness of the proposed method is demonstrated by the experimental results.
Keywords : Face detection, Gabor wavelet, Feed Forward Neural Network (FFNN) classifier, Multilayer Perceptron.
382-390
PDF
16
Heterogeneous Mesh Network for LTE Data Offloading using Specialized Nodes
-Jerin M James,Dr. Milind Thomas Themalil
Abstract
Congestion in LTE and other cellular networks can be reduced by cell splitting and by buying more spectrum. But due to the heavy licensing fee and the cost of setting up new base stations, service providers have started to resort to the unlicensed spectrum, particularly the ISM band to offload some of the data from the LTE network locally, in areas of heavy user base. There does not exist a long-term and standard solution although some efforts have been made to enable cooperation between the different networks like LTE and IEEE 802.11. Thus, a new architecture and accompanying routing protocol is proposed to enable a tighter integration between different networks to improve QoS and overall performance of the combined heterogeneous network. The performance improvement is 2-3 times more than LTE and Wi-Fi networks when used independently, even when twice the bandwidth is used by the component networks alone.
391-395
PDF
17
A Review on Numerical Protection Relays for Electrical Substation
-Atul L. Suryawanshi,Dr.Gayatri M. Phade
Abstract
In view of the global scenario, electricity has become a prime need of human being, ranging from household to industrial needs. The plexus of electricity from generation to distribution is enormous. One of the most complex disciplines in electrical engineering is power system protection which not only requires proper understanding of the different components of a power system but also a good knowledge and analysis of the abnormal circumstances and failures that can occur in any element of a power system. The research of load flow analysis, fault level calculations and figuring out the maintenance and repairs is very intricate because one has to mull over the line routes and relay settings are to be decided based on the worst case in the fault prone area. It also becomes obligatory to conduct several trials so as to calculate the time multiplier scale of numerical protection relays which ensures that the minimum differentiation margin is found between main protection relay and its backup relays. This can only be possible by implementing a protection algorithm using an embedded system. This paper sheds light on the evolution of protective relays since the onset of electrical energy & different techniques of building a protection. We try also to foresee the future prospects and trends in this area, considering the advantages of latest developing systems in Power System Protection and Substation Automation domain.
Keywords— electricity; failure; power systems; numerical protection relays;substation automation
396-399
PDF
18
Enhanced the Data Security of Cloud Based Services Using RSA and DES Encryption Algorithm
-Himanshu Gupta,Mahesh Motwani,Sanjay Silakari
Abstract
The data security play an important role in cloud based services. The cloud based services used by various social networking website such as Facebook, twitter, Instagram and many more android based applications faces the threats of security on both sides user as well as server side. For the enhancement of security strength there are various cryptography algorithm such as symmetric and asymmetric algorithm. In this paper we use RSA and DES cryptography algorithm for user data encryption on cloud environments. We have used cloudsim simulator and used different scenario of load for analysis on both RSA and DES encryption algorithms.
Keywords: - Cloud Computing, Cryptography, RSA, DES, Cloudsim
400-404
PDF
19
Early Detection of Cholesteatoma using Imaginary Techniques-a Noval Approach
-Nupur Tewari,Kirti Somani,Khandakar Faridar Rahman,Saurabh Mukherjee
Abstract
In pathological terms cholesteatoma can be stated as a collection of squamous epithelium (i.e. abnormal skin growth) behind the ear drum in the middle ear. In this paper, we have studied the various imaging techniques that help in the diagnoses of cholesteatoma and after carefully analyzing them; we have developed two algorithms to overcome the difficulties faced during our research period and have also done a comparative study of the two proposed algorithms. Experiments are performed on the secondary image data set available for middle ear cholesteatoma using MATLAB Tool. Experiment results show the performance measure of the two algorithms.
Keywords — Medical imaging, Computed Tomography, Magnetic resonance imaging, Gray Level Co-occurrence Matrix, Mathematical morphology, Image Segmentation
405-411
PDF
20
A Heuristic approach for securing data in cloud environment based on Text Classification and Machine Learning techniques
-N.A.Samasth,Deepak.G,Harish Kumar N,Rajath.R.Kulkarni,Pavan.S. Naik
Abstract
Cloud computing is a general term for the delivery of hosted services over the internet. Cloud primarily refers to the saving of customers' data to an off-site storage system that is preserved by a third party. Emerge of cloud computing has empower the possibility of doing complex calculations, mass storage and device operations to the cloud in a powerful way. But the rising security problems have resisted the organizations from connecting with cloud computing completely hence security risks have appeared as the main disadvantages of cloud computing. The existing solutions encrypt all data using the same key size without taking into consideration the confidentiality level of data which in turn will increase the cost and processing time. While addressing the security issues the first important step is to classify what data which requires higher security. To achieve this, a machine learning approach has been applied. The proposed cloud model minimizes the overhead and processing time needed to secure data through using different security mechanisms with variable key sizes to provide the appropriate confidentiality level required for the data.
412-416
PDF
 
 
IJCTA © Copyrights 2015| All Rights Reserved.

This work is licensed under a Creative Commons Attribution 2.5 India License.