Sanjeev Sofat

Detection of edges in color images: a review and evaluative comparison of state-of-the-art techniques

We present an evaluative review of various edge detection techniques for color images that have been proposed in the last two decades. The statistics shows that color images contain 10% additional edge information as compared to their gray scale counterparts. This additional information is crucial for certain computer vision tasks. Although, several reviews of the work on gray scale edge detection are available, color edge detection has few. The latest review on color edge detection is presented by Koschan and Abidi in 2005. Much advancement in color edge detection has been made since then, and thus, a thorough review of state-of-art color edge techniques is much needed. The paper makes a review and evaluation of various color edge detection techniques to quantify their accuracy and robustness against noise. It is found that Minimum Vector Dispersion (MVD) edge detector has the best edge detection accuracy and Robust Color Morphological Gradient-Median-Mean (RCMG-MM) edge detector has highest robustness against the noise.

Deployment and evaluation of IEEE 802.11 based wireless mesh networks in campus environment

Wireless Mesh Networks (WMNs) can be considered as hybrid between wireless infrastructure (WLAN) and ad-hoc networks (MANET), with mesh points providing flexibility in building & expanding the network, allowing automatic discovery of neighboring nodes, increased reliability and redundancy. In this paper we discuss as to how WMNs can be practically deployed to support wireless multihop communications in a campuswide area. To this aim, we have deployed a real WMN at PEC University of Technology campus utilizing state-of-the-art technology and analyzed the performance of this architecture when supporting multihop heterogeneous traffic. Currently the network is being used to provide services to the residential areas of the campus.

Template and Database Security in Biometrics Systems: A Challenging Task

Security is a very important aspect in the biometric system. There are number attacks and there remedial solutions discussed in the literature on different modules of biometrics system and communication links among them. But still the researchers are not able to secure every module of a biometric system against these attacks. Template and database are the very important parts of biometric systems and attacker mostly attack on template and database of biometric system so securing them is a very crucial issue these days. In this research paper our focus is on template and data base security in biometrics system and we develop a system to encrypt and decrypt the biometric image using helper data of a fingerprint and password to make it secure so that even if someone gains access to the encrypted image stored in the database he will not able to reproduce the original image from it and it will be useless for him.

Techniques to Detect Spammers in Twitter- A Survey

With the rapid growth of social networking sites for communicating, sharing, storing and managing significant information, it is attracting cybercriminals who misuse the Web to exploit vulnerabilities for their illicit benefits. Forged online accounts crack up every day. Impersonators, phishers, scammers and spammers crop up all the time in Online Social Networks (OSNs), and are harder to identify. Spammers are the users who send unsolicited messages to a large audience with the intention of advertising some product or to lure victims to click on malicious links or infecting user’s system just for the purpose of making money. A lot of research has been done to detect spam profiles in OSNs. In this paper we have reviewed the existing techniques for detecting spam users in Twitter social network. Features for the detection of spammers could be user based or content based or both. Current study provides an overview of the methods, features used, detection rate and their limitations (if any) for detecting spam profiles mainly in Twitter.

Detecting Malicious Users in Twitter using Classifiers

The web has become a vital global platform that binds together almost all daily activities like communication, sharing, and collaboration. Impersonators, phishers, scammers and spammers crop up all the time in Online Social Networks (OSNs), and are even harder to identify. People in the public eyes like politicians, celebrities, sports persons, media persons and other public figures with huge followings are particularly vulnerable to this type of attacks. The main objective in this work is to identify those forged users who harm genuine ones, jeopardize the identity and hence the security and privacy of users. In this paper a framework for the detection of malicious users, non-malicious users and celebrities has been developed by using an attribute set for user classification based on user characteristics. For the purpose of detecting malicious users, non-malicious users and celebrities, a crawler has been developed for Twitter and data of around 22K users have been collected from publicly available information. Data of around 7,500 users have been used for training and testing purpose in Weka for classification of users. 5 classifiers have been used and compared on the basis of performance metrics like precision, recall, F-measure and accuracy. RandomForest outperforms all the classifiers with 99.8% accuracy.

Deauthentication/Disassociation Attack: Implementation and Security in Wireless Mesh Networks

M esh Networks have emerged as a widely deployed, new paradigm with improved performance and reliability. M esh Networks offer ubiquitous network connectivity along with better flexibility and adaptability features. Despite of these benefits, Wireless M esh Networks are vulnerable to attacks due to the absence of trusted central authority and the unprotected nature of the management frames. This security breach leads to the spoofing of legitimate clients information. Thus facilitating the launch of dos attacks on the behalf of the legitimate identity holders .The influence of DOS attacks is highly intense, because complete network resources have been consumed by the attacker after launch of the attack. Consequently, it leads to deterioration of network performance thus halting the communication. Therefore, security is a major concern that needs to be dealt with to alleviate the effect of these attacks . So that the deterioration and disruption caused by these attacks to the network performance has been thwarted. In this paper we have implemented the dos attacks on the real wireless mesh test bed and analyzed their impact on the network performance and proposed a security algorithm for the detection of these attacks.

Web hypermedia content management system effort estimation model

This study aims at creation of a well defined estimation model which can be used to estimate the effort required for designing and developing the web hypermedia content management systems. The data from the different content management system projects are studied and the linear regression approach is used to finalize the model. This model also provides guidelines to calculate phase wise distribution of effort. The model is designed to help project manager to estimate effort at the very early stage of requirement analysis. A set of questionnaire is used to estimate the complexity of the project, which has to be filled after completing the initial requirement analysis. Final effort is estimated using the project size and the different adjustment factors. For better calculation of these adjustments factors, these are categorized into three categories based on their characteristics such as Production and General system characteristics. This model is proposed to be used differently for the different types of projects. These projects are categorized based on their size and total/build effort ratio. The size of the project is estimated by using the modified object point analysis approach. The estimated effort is further phase wise distributed for better scheduling of the project. Another questionnaire is used to refine the model and it has to be filled by the project managers after completing the project. The proposed model is validated by studying twelve completed projects taken from industry and seventy different projects completed by the students. The proposed model shows a great improvement as compared to the earlier models used in effort estimation of CMS projects.

Integrated Framework for Classification of Malwares

Malware is one of the most terrible and major security threats facing the Internet today. It is evolving, becoming more sophisticated and using new ways to target computers and mobile devices. The traditional defences like antivirus softwares typically rely on signature based methods and are unable to detect previously unseen malwares. Machine learning approaches have been adopted to classify malwares based on the features extracted using static or dynamic analysis. Both type of malware analysis have their pros and cons. In this paper, we propose a classification framework which uses integration of both static and dynamic features for distinguishing malwares from clean files. A real world corpus of recent malwares is used to validate the proposed approach. The experimental results, based on a dataset of 998 malwares and 428 cleanware files provide an accuracy of 99.58% indicating that the hybrid approach enhances the accuracy rate of malware detection and classification over the results obtained when these features are considered separately.

Classification of PE Files using Static Analysis

Malware is one of the most terrible and major security threats facing the Internet today. Anti-malware vendors are challenged to identify, classify and counter new malwares due to the obfuscation techniques being used by malware authors. In this paper, we present a simple, fast and scalable method of differentiating malwares from cleanwares on the basis of features extracted from Windows PE files. The features used in this work are Suspicious Section Count and Function Call Frequency. After automatically extracting features of executables, we use machine learning algorithms available in WEKA library to classify them into malwares and cleanwares. Our experimental results provide an accuracy of over 98% for a data set of 3,087 executable files including 2,460 malwares and 627 cleanwares. Based on the results obtained, we conclude that the Function Call Frequency feature derived from the static analysis method plays a significant role in distinguishing malware files from benign ones.

A Novel Color Coherence Vector Based Obstacle Detection Algorithm for Textured Environments

 Abstract—This paper presents a novel obstacle detection algorithm that makes use of color information and color coherence vectors for robust obstacle detection. The algorithm makes use of color cue to classify a pixel in an image as an obstacle or a path. Color is one of the prominent image features. Color information is readily available as input from a color camera. Our algorithm makes use of coherence vectors for representation and matching instead of histograms. A color histogram provides no spatial information. It merely describes the color information present in an image. Color coherence vectors represent pixels as either coherent or incoherent. Color coherence vectors prevent coherent pixels from getting matched with incoherent pixels. The color histogram cannot make such fine distinction. The algorithm is tested with a challenging indoor and outdoor image set. Test results show that our algorithm performs better than available color based obstacle detection approaches.