Volume 1, Issue 9, 2011

By | August 9, 2018

A Hybrid Method for Enhancement of Plant Leaf Recognition 

N.Valliammal, S.N.Geethalakshmi

Abstract: This paper focuses on the preprocessing technique for CAP-LR   (Computer aided plant classification through leaf recognition). Pre   processing is the basic step to reconstruct the image with some useful   feature. This technique is essential for the enhancement of leaf images   which increases the efficiency of the subsequent tasks of the leaf   recognition system. In this paper, an hybrid approach is proposed which is a   combination of contrast stretching and adaptive thresholding that   simultaneously adjusts the intensity level of leaf images using boundaries is   developed. The validation of proposed system is carried out based on the   defined parameter matrices. The experimental results shows that the   proposed method proves efficient when compared to other traditional   methods. 

Keywords : image enhancement; Histogram equalization; Contrast   stretching; intensity adjustment; Adaptive Thresholding; Median filter;   wavelet filter. 


Effect of Pressure Variation on the Fingerprint Minutiae Based Feature Vector Matching 

R. N. Verma, D.S.Chauhan

Abstract: Fingerprint verification of online fingerprint images has become a   very actively studied field in biometric systems. This paper investigates the major challenges in fingerprint recognition if the   images are captured at varying pressure. Extracting minutiae  from  fingerprint  images  is one of  the  most  important  steps  in   automatic  fingerprint verification  and  classification. Minutiae are local discontinuities in the fingerprint images, mainly ridge termination   and bifurcations. In this paper we proposed an algorithm   m-margin to fill the broken curves on a fingerprint. Our proposed algorithm   eliminates false minutiae that connect broken curves in    fingerprint. The method is initialized with the captured image using local   histogram equalization, FFT, image binarization, thinning, skeleton refinement, end-points, filtering and matching. The  images  achieved from these steps are passed through proposed mmargin algorithm. This method show some improvement in the minutiae extraction process in   terms of either efficiency and/or time   required. The quality of the fingerprint images greatly affects the   performance of the minutiae extraction.  

Keywords : biometrics; minutiae; ridge; valley; FRR; FAR; skeletonization;   m-margin. 


Cooperative Machine Learning Method 

Adnan M. Al-Khatib

Abstract: Cooperative learning refers to an approach where one or more   team of learners work together towards reaching a better knowledge or   understanding of a specified task. The purpose of this paper is to use this   approach to describe a proposal for designing and building a cooperative   machine learning system (Multi-Learning system) that contains two or more   machine learners that cooperate together.   The objective of the proposal in this paper is to improve the effectiveness   of the machine learning systems. 

Keywords : Cooperative Learning; Machine Learning; inductive learning;   Multi-Learning system; Knowledge base; Artificial Intelligence.  


Cyclic model application for evaluation of a Dynamic Web   Environment

Sonal Chawla, Ritika Bansal

Abstract: Dynamic web environment in the form of E-learning systems are   increasingly being adopted by educational   organizations. This E-learning system is taking important place in formal   and less formal educational setting. Such an environment   enhances the conventional system of education by utilizing the digital   resources comprising of learning objects as delivery mode for imparting instructions. This research paper, thus, focuses on the creation   and evaluation of a Dynamic Web Environment and has fourfold objective. Firstly, it focuses on dynamic web environment,  learning objects and their relation in development of dynamic   web environment. The paper then pre-tests the perception of students for   such an environment Secondly, it discusses the creation of dynamic web environment with special reference to the research study   conducted by creating  a demonstrative dynamic web   environment called E-campus. The tools and technologies used in the   development of this dynamic web environment are also   highlighted. Thirdly, the paper post-tests the created E-Campus environment.  And finally, the  paper ends with analysis and conclusions drawn for these research studies.   

Keywords : Interactive learning environments; Improving classroom   teaching; Lifelong learning;  Cooperative/Collaborative   learning; Teaching/learning strategies. 


Study of Effect of Rate on Performance of Cross Layer Based Inrusion Detection for WLAN 

Ravneet Kaur

Abstract: The advent of wireless LAN is a onelcome sign in terms of ease of   deployment and reconfiguration. Hoonever, security is an alarming concern, as everything being transmitted is available in the   air. The presernt paper deals with Study of effect of rate on performance of cross layer based intrusionn detection for WLAN reflects   the significance of cross layer technique in detecting intruder on WLAN. In cross layer based intrusions detection the decision is based on the combine oneight value of two or more layer. So the decision is not based on single layer, it will reduce false positive rate. Two different layers, physical and MAChave been used in the present study and the results have been compared with existing techniques.

Keywords- Receiver Signal Strength (RSS); Time Taken for RTS-CTS Handshake (TT); Radio Frequency (RF).

Leave a Reply

Your email address will not be published. Required fields are marked *