Notice Board :





Volume II Issue III

Author Name
R Stanica, K A Ali
Year Of Publication
2010
Volume and Issue
Volume 2 Issue 3
Abstract
Digital watermarking has been an area of constant research due to the internet pirates who use the internet beyond knowledge and entertainment. Inexpensive distribution and transmission of multimedia data has led to the need of effective copyright protection tools and digital watermarking is one of them. This paper is an attempt to develop a secure stereo image watermarking system which can protect multimedia data (stereo image in our case) from unauthorized users and establish copyright in case of multiple ownership claims. The host image has been firstly encrypted using the famous double random phase encoding technique instead of the most commonly used scrambling algorithms that change the pixel positions only and not the gray values. Watermarks (disparity image and another containing owner’s identity) are then embedded in different sub-bands of discrete wavelet transform to establish copyright. Embedding strengths of both the watermarks has been optimized using genetic algorithm. Genetic algorithm, an excellent optimization technique prove their superiority in watermarking by maximizing both the requirements of watermarking i.e. imperceptibility and robustness. The results alone are sufficient to prove the superiority of the algorithm.
PaperID
2010/EUSRM/02/03/1004

Author Name
A abdeld, N U Rehman
Year Of Publication
2010
Volume and Issue
Volume 2 Issue 3
Abstract
Numerous real life applications can use character recognition system for reliability, automation, speed and consistency. From industry to home, there are so many situations where character recognition system can play vital role. In this paper, we have developed handwritten character recognition system which various feature extraction tool. The algorithms use various features such as boundary profiles in all directions, horizontal and vertical projection histograms and chain code of alphabet skeleton. The segmentation of alphabet is done by using morphological operations. The local and global features obtained from the segmented alphabet are feed to the neural network structure in order to classify the input alphabet. For analysing the impact of multiple feature extraction we use different training samples of handwriting The use of the character recognition is well known and its use is growing in the sector of education.
PaperID
2010/EUSRM/02/03/1011

Paper Title
Author Name
H J F Lyim
Year Of Publication
2010
Volume and Issue
Volume 2 Issue 3
Abstract
A QR Code is a specific matrix barcode (or two-dimensional code), readable by dedicated QR barcode readers and camera phones. The code consists of black modules arranged in a square pattern on a white background. The information encoded can be text, URL or other data. QR codes storing addresses and URLs may appear in magazines, on signs, buses, business cards, or on just about any object about which users might need information. Users with a camera phone equipped with the correct reader application can scan the image of the QR Code to display text, contact information, connect to a wireless network, or open a web page in the phone's browser. This act of linking from physical world objects is known as a hard link or physical world hyperlinks. Google's mobile Android operating system supports the use of QR codes by natively including the barcode scanner (ZXing) on some models, and the browser supports URI redirection, which allows QR CODES to send metadata to existing applications on the device. Nokia's Symbian operating system is also provided with a barcode scanner, which is able to read QR codes, while mbarcode is a QR code reader for the Maemo operating system. In the Apple iOS, a QR code reader is not natively included, but over 50 free Apps are available with reader and metadata browser URI redirection capability.
PaperID
2010/EUSRM/02/03/1015

Author Name
O Arias-Carrión, Y WEI
Year Of Publication
2010
Volume and Issue
Volume 2 Issue 3
Abstract
Research in the field of computer and network science demands for tools and methodology to test their security effectively. Intrusion Detection System is used to perform the same with a fact that an intruder’s behavior will be noticeably different from that of a legitimate user and would exploit security vulnerabilities. IDS have thousands of alerts per day; some are mistakenly triggered by begin events. This make it extremely difficult to correctly identify alerts related to attack. In this paper, we propose neural network based method for network intrusion detection. These technique are applied to the KDD Cup 98 data set .In addition, a comparative analysis shows the advantage of Unsupervised Learning techniques over clustering-based Methods in identifying new or unseen attack.
PaperID
2010/EUSRM/02/03/1019