|
|
|
|
Home>Products>Journals and Reviews>I.Re.Co.S.>Latest issue International Review on Computers and Software Vol. 6 N. 6-- Part A International Review on Computers and Software Vol. 6 N. 6-- Part B
|
International Review on Computers and Software - Papers- Part A
Recommendation
Systems through Semantic Web Usage Mining
by Mehrdad
Jalali, Norwati Mustapha
Abstract - Internet users are drowned in all kind of available information. However, only a tiny part of it is usually relevant to their preferences. Web usage mining which extracts knowledge for usage and clickstream data has become the subject of exhaustive research, as its potential for Web-based personalized services, prediction of user near future intentions, adaptive Web sites, and customer profiling are recognized. Moreover, semantic web aims to enrich the WWW by machine processable information which supports the user in his tasks. Semantic clickstream mining which integrates semantic in Web usage mining processes aims to improve the quality of the Web usage mining systems. Given the primarily syntactical nature of data Web usage mining operates on, the discovery of meaning is impossible based on these data only. Therefore, formalizations of the semantics of Web resources and navigation behavior are increasingly being used. In this paper, we discuss the interplay of the Semantic Web with Web usage mining and also we give an overview of where the two research areas meet today. Moreover a proposed framework to integrate semantic Web and Web usage mining discuss in the rest of the paper.
Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved
Keywords: Web Usage Mining, Clickstream Mining, Semantic Web, Semantic Web Usage Mining.
Prototyping
Mobile Graphical Authentication
by
Martin Mihajlov, Borka Jerman Blažič, Dragan Tevdovski
Abstract - As users can only make informed choices when the proposals being discussed are meaningful to them, enabling users to envision and make sense of those proposals is an essential element of all approaches to system design. The focus of this study is the conceptualization of a graphical authentication mechanism appropriate for ubiquitous environments. Therefore, the experiment presented in this paper uses the paper prototyping method is used to effectively simulate graphical authentication and distinguish the user preference between two system approaches in ubiquitous environments. Complementary, both a high fidelity paper and a mobile prototype is used in order to evaluate popular graphical authentication concepts which are based on separate cognitive functions: recall and recognition. This experiment uses a between group design where each participant evaluates a prototype for both authentication concepts in the same medium. The results of the study demonstrate that recognition-based graphical authentication mechanisms are more suitable for ubiquitous environments than recall-based systems.
Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved
Keywords: Graphical Authentication, Mobile Prototyping, System Evaluation.
Some
Open Problems in Multimedia Digital Fingerprinting
by Song Tang, Li Liu
Abstract - As an important digital copyright protection technology developed in recent years, digital fingerprinting achieves piracy tracing by embedding a unique serial number called fingerprint into each distributed multimedia copy invisibly. When a pirated copy is found somewhere, the embedded fingerprint can uniquely identify the source of the leakage. Although there are many studies on the digital fingerprinting, there are still some problems have not been satisfactorily resolved. There are many open problems regarding the digital fingerprinting technology, both theoretical and practical nature. In this paper, we analyze and discuss some of the major open problems in the field of digital fingerprinting, which are the fingerprint embedding, the collusion attack and the effective distribution of the differently marked copies.
Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved
Keywords: Multimedia Copyright Protection, Digital Fingerprinting, Information Hiding, Secure Content Distribution, Collusion Attack.
QoS
Correction in IMS Networks
by .
Errais, B. Raouyane, M. Bellafkih, M. Ramdani
Abstract - The network management, especially monitoring and correction of QoS degradation, is becoming an increasingly difficult task. Indeed the network size evolution and diversity of deployed technologies in new architectures. Particularly in IMS networks, which the management tasks and resource configuration is complex operation. In this paper, a new architecture for management and control IMS networks, based on eTOM specifications. The architecture aims to move toward an independent system for monitoring and correction of QoS degradation automatically in real time.
Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved
Keywords: IP Multimedia Subsystem (IMS), Enhanced Telecom Operation Management (eTOM), Network Management, QoS Control.
A
Strawberry Disease Image Retrieval Method Inosculating Color and
Textural Features
by Jian Song
Abstract – A strawberry disease image retrieval method inosculating color and textural features is proposed in order to provide farmers with visualized guidance for disease and insect control and overcome the inaccuracy of the diagnosis expert system for vegatable deseases which relies on words to provide information. A detailed analysis is mdade on the color features of the strawberry disease images, which are preprocessed in HSV color space which is in accord with visual features of human eyes. 4×4 image feature matrix is constructed when the image mean value, variance, measure of skewness, kurtosis, and energy are extracted as the color feature value while the image contrast, texture consistency and the relevance of entropy and grayscale as the textural feature value. After they are normalized by Gaussian normalization method, Mahalanobis Distance is adopted for similarity measurement. A strawberry disease image retrieval method inosculating color and textural features is developed with C++ programimg under Visual C++6.0 development environment. It is indicated by test results that this retrieval methode has a preferable recognition effect with a Precision Ratio of 65% and a recall ratio of 83%. When this algorithm is applied to the diagnosis expert system for strawberry diseases, its robustness will be greatly enhanced to meet the requirments of disease diagnosis.
Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved
Keywords: Strawberry, Disease Image, Contentbased Image Retrieval, Color Features, Textural Features.
A
Robust Audio Watermarking Scheme Based on SVD and MDCT
by Ping Su,
Haidong Shi
Abstract – A lot of scholars have successfully applied SVD (Singular Value Decomposition) in watermarking algorithm of digital image. But the application research of SVD in audio digital watermarking only appears in recent two years; one major reason is that the audio signal is one-dimensional signal in time domain, which can not directly carry out SVD decomposition for it. At present, classic audio coding compression standards (such as MP3, AAC, AC3, etc) mostly adopt MDCT (Modified Discrete Cosine Transform) to complete the time frequency transform for audio data. A robust audio watermarking algorithm has been proposed in this paper. First, the MDCT coefficient matrix of original audio is built. Then the watermark is embedded by SVD and adaptive strategy is applied during embedding. The experimental results show that the proposed algorithm has good imperceptibility and robustness for low-pass filtering, noise adding and loss compression.
Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved
Keywords: Singular Value Decomposition, Modified Discrete Cosine Transform, Audio Watermarking, Time Frequency Transform.
Recognition
of Handwritten Arabic Characters by Using Reduction Techniques
by
Salah M. Al-Saleh, Salameh A. Mjlae, Salim A. Alkhawaldeh
Abstract -
Instance-based learning (IBL) is one of the simplest classification
methods that is used for more sophisticated learning techniques such as
neural networks. However, IBL has large number of stored instances which
leads to the increase in the classification execution time. Therefore, a
number of reduction techniques have been developed to overcome this
problem. Unfortunately, these reduction techniques were not applied for
Handwritten Arabic character recognition. In this paper, we apply four
reduction techniques for Handwritten Arabic character recognition. The
applied reduction method are Condensed Nearest Neighbour (CNN), Edited
Nearest Neighbour (ENN), Instance-Based Learning Algorithm 2 (IB2) and
Instance-Based Learning Algorithm 3 (IB3). In addition, a technique that
performs a pre-processing of the handwritten Arabic characters patterns
is proposed. A resulted data set with a small number of instances
related to these patterns is presented. Numerical results presented that
the size of the resulted training set is much small than that of the KNN
technique with acceptable classification accuracy.
Copyright © 2011 Praise Worthy Prize S.r.l. -
All rights reserved
Keywords: IBL, Image Pre-processing, Pattern Recognition, Reduction Methods .
SVM-kNN
Fusion in Vocabulary Tree Method for Specific Object Recognition
by
Amir Azizi, Sattar Mirzakuchaki
Abstract - Object
recognition is a crucial task in computer vision. In recent years,
popularity of object recognition techniques which are based on local
features has increased among computer vision researchers.This is due to
the robustness of these approaches to real scenes challenges such as
viewpoint changes, occlusion, and illumination variations. Vocabulary
tree is one of the object recognition techniquesbased on local features
and is used in robotics due to its speed. Thus increasing the accuracy
of this method becomes important. In this paper, in addition to
evaluation of vocabulary tree’s accuracy for specific object recognition
on a difficult dataset we show that the fusion of SVM and kNN
classifiers in vocabulary tree leads to an increase in accuracy.
Copyright © 2011 Praise Worthy Prize S.r.l. -
All rights reserved
Keywords: Local Features, Specific Object Recognition, SVM-kNN Fusion, Vocabulary Tree.
Energy
Efficient Routing Protocols for Wireless Sensor Networks: a Surveyby K S Shivaprakasha, Muralidhar Kulkarni
Abstract - Wireless Sensor Networks (WSNs) have become one of the emerging trends of the modern communication systems. They find their applications in various fields like habitat monitoring, home automation, environment monitoring, battle filed environment etc. WSNs are different from Mobile Adhoc Networks in the perspective of energy awareness, adaptive communication patterns and the routing algorithms. As the sensor devices are powered by batteries, which cannot be recharged often, the power awareness is one of the major requirements in WSNs. Many energy aware routing protocols have been proposed in the literature. In this survey, an attempt has been made to summarize the various energy aware routing protocols available in the literature and also a comparative analysis of these has been made considering various network parameters like the delay, routing overhead, QoS, type of routing protocol etc.
Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved
Keywords: Wireless Sensor Networks (WSN), Base Station (BS), Cluster Head (CH), Medium Access Control (MAC), Region Head (RH).
Efficiency
of MDB Routing Algorithm over DB Routing Algorithm in Point-Point Networks
by S.Anuradha, G.
Raghu Ram, T. Bhaskara Reddy, J.Chakradhar
Abstract - The goal of this paper is to investigate to control the congestion in the Communication Networks. The study is based on use of an ant algorithm operating upon a dynamic problem domain, that is, a domain that changes as a function over time. Specifically, discussed about the use of an ant-inspired graph-based general- purpose algorithm metaheuristic named Ant Colony Optimization as the basis of the method of implementation of a routing algorithm in a packet- switched point-to-point network (such as the Internet). Ant-inspired algorithms have the capability of finding short paths in graphs, and show an inherent adaptability that could be utilized to solve dynamic problems such as routing with node balancing in a network. Here a new general purpose heuristic algorithm named (MDB) Modified Depth-Breadth routing algorithm is defined and its performance was compared with (DB) Depth-Breadth routing algorithm and can be prove more efficient than DB routing algorithm.
Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved
Keywords: ACO, DB-routing, MDB-routing, Congestion Control.
Real
Time "2CRT" Architecture and Platform for the Experimentation of
Telecommunication Terminals According to the Manhattan Mobility
by
M. ElBakkali, H. Medromi
Abstract – In the field of technology the verification of the robustness of a system is very important. This system verification compares the actual with the origin bearing in mind that any anomaly could lead to a catastrophe. The control could be in different forms: either measure, or test the first control is done to measure the physical parameters in order to confirm system reliability as it is the case of « 2CRM » (Connection, Control, Recognition and Measure) [1]. The second control enables a remote experimentation to certify credibility of the terminal. Our article presents a new architecture of control « 2CRT » (Connection, Control, Recognition and Test). Its newness is based on the distribution of actions as it is based on multi-agent systems. The control is done remotely via the web using Real Time constraints. It meets to the functionalities of distribution, cooperation and adaptation to the changes in user behavior, environments and materials. Notably, the approach includes the test control mode of telecommunication terminals. The solution is based upon the Manhattan mobility with the possibility of integrating other topologies. This experimentation is done via the web by matrix configurations using FIFO (First In First Out) data structures. To further embody our architecture, we perfected it in the OpenAirInterface project of the Eurecom laboratory in France. Our control platform was developed using the unified process « RUP » (Rational Unified Process) [2] with the use of the UML modelization to well illustrate the system. It is implemented by an open source distribution (Java and J2ee running on Eclipse).
Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved
Keywords: Measure Control, Test Control, Distributed Systems, Real Time, 2CRM, 2CRT, Multi-Agent Systems, Physical Parameters, Labyrinth, Telecommunication Terminals, Manhattan Mobility, Matrix Configurations, FIFO, Openairinterface, RUP, UML, J2ee.
A
Bandwidth Request Competition Mechanism in WiMAX System
by
Jianbo Ye, Qingbo Zhang
Abstract - Bandwidth
competition resolution mechanism is a key technology of WiMAX. According
to WiMAX agreement, Bandwidth competition resolution mechanism adopts
the binary exponential return algorithm generally. But this algorithm
will reduce the loss rate of bandwidth request and increase the number
of SS and the loss rate bandwidth request as the initial return window
increased. This paper proposes a bandwidth competition system based on a
request to wait in WiMAX system. The proposed algorithm limits the
number of bandwidth request sent in one frame, that is to say each user
does not send new bandwidth request immediately after a request is
successfully sent, but rather to wait for several frames before sending.
The simulation results show that the proposed algorithm not only the
method is simple, but also it can effectively reduce the loss rate and
delay of bandwidth request. So, the proposed algorithm has practical
engineering meaning.
Copyright © 2011 Praise Worthy Prize S.r.l. -
All rights reserved
Keywords: WiMAX, Bandwidth Competition, A Request to Wait, Loss Rate, Delay.
An
Inter-cluster Cooperative Nodes Selection Scheme Based on Blind Channel
Estimation
by
Xiaoqiang Zhong, Baiqing Zhou
Abstract - An
inter-cluster cooperative node selection scheme with low energy-cost is
proposed for the energy –constrained wireless sensor networks. Pre-set
an energy threshold first and determine a cooperative node set. Then
estimate information of channels between inner-cluster cooperative nodes
and cluster head nodes with a blind channel estimation algorithm that
requires small amount data. The optimum inner-cluster nodes are selected
as the cooperative nodes of the cluster head nodes by considering
comprehensively channel states and residual energy. The blind channel
estimation is then turn into the unconstrained optimization which is
then resolved by quadratic programming iterative weighted method.
Theoretical analysis, mathematical models, and simulation results show
that the new algorithm can obtain channels information by mean of a few
data.
Copyright © 2011 Praise Worthy Prize S.r.l. -
All rights reserved
Keywords: Wireless Sensor Networks, Cooperative Communication, Blind Channel Estimation, Quadratic Programming.
A
Suitable Architecture and Deployment Considerations for Shallow Water Acoustic
Sensor Networks
by
V. Hadidi, R. Javidan M. Keshtgary, A. Hadidi
Abstract - Nodes of
an underwater Acoustic Sensor Network (UASN) are connected together
through acoustic sound that propagates in the water better than
electromagnetic waves. They are envisioned to enable applications for
real-time underwater monitoring and data collection. Moreover, unmanned
or autonomous underwater vehicles (UUVs, AUVs), equipped with sensors,
enable the exploration of natural undersea resources and gathering of
scientific data in collaborative monitoring missions. In this paper,
several fundamental key aspects of underwater acoustic communication are
investigated, Different architectures for two-dimensional and
three-dimensional underwater sensor networks are discussed,andthe main
challenges for the development of efficient networking solutions are
presented.Next, we propose a new multipath scheme for shallow water,
which can guarantee certain end-to-end packet error rate while achieving
a good balance between the overall energy efficiency and the end-to-end
packet delay. Finally an appropriate localization algorithm forshallow
waters is proposed. The simulations resultson prototype data show the
effectiveness of the proposed architecture and localization algorithm.
Copyright © 2011 Praise Worthy Prize S.r.l. -
All rights reserved
Keywords: Acoustic Communications, Routing Protocols, Architecture of Underwater Acoustic Sensor Networks, Localization Algorithm.
Creating
Barcode Using Fractal Theory
by K.
Kiani, E. Kosari, M. R. Simard
Abstract - This article introduces and explains a new method of 2D coding based on L-Systems Fractal Hypothesis which has a better security and capacity for coding. The coding, image processing and decoding of two dimensional codes by Hopfield Neural network or a simple decision function are studied step by step in this article. Different case studies for a new barcode shape show that the current approach provides more reliable method in the early stage of design process.
Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved
Keywords: Barcode, Fractal, Image Processing, L-System .
Model
Proving of Urban Traffic Control Using Neuro Petri Nets and Fuzzy Logic
by
Rishi Asthana, Nilu Jyothi Ahuja, Manuj Darbari
Abstract - The
development of control systems to handle the congestion at intersection
in urban traffic is a critical research issue. Petri Nets and Fuzzy
Logic have played vital role in the development of such systems. Petri
nets, a mathematical modeling tool, makes graphical modeling, simulation
and real time control modeling, more functional. Fuzzy logic deals with
uncertainties in the environment to make the systems more realistic. In
this paper, a real time traffic control model has proposed and
implemented in MATLAB. Several results have been discussed and found
satisfactory.
Copyright © 2011 Praise Worthy Prize S.r.l. -
All rights reserved
Keywords: Fuzzy Logic, Petri Nets, Neuro Petri Nets, Membership Functions (MFs).
A
New Evaluation Strategy Based on SVM for Supply Chain Finance
by Wenfang Sun,
Jindong Zhang, Xilong Qu
Abstract - The concept of supply chain finance is introduced, The SAW(Self-avoiding random walk) model characterizing fractal dimension of biological macro molecular fractal network is analyzed. Support vector machine algorithm has been adopted to become an evaluation strategy for supply chain finance. The evaluation and choice for kernel function is very important. In this paper, a quantity estimate is proposed though empirical risk and confidence interval based on the structural risk theory. The results of simulation experiments are shown, and the feasibility and effectiveness of this method is proved.
Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved
Keywords: Supply Chain Finance; Kernel Function; Support Vector Machine.
Leveraging
Historical Assessment Records for Constructing Concept Maps
by M. Al-Sarem, M.
Bellafkih, M. Ramdani
Abstract - In recent years, researchers have proposed different approaches for developing adaptive learning systems based on learning behaviors of learners during their interaction with e-learning systems. For achieving the adaptive learning, a predefined concept map of a course is often used to provide adaptive learning guidance for learners. However, it is difficult and time consuming to create the concept map of a course. In this paper, we apply Data mining techniques to constructing concept maps based on the historical testing records. The paper, first, provides a common basis for the analysis of automatic concept maps building approaches meeting in literatures. Then, the general process of automatically constructing concept maps using historical assessment records is proposed.
Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved
Keywords: Association Rules, Concept Map, Domain Model Representation, Educational Data Mining, Formal Student Model.
Object
Oriented Database Applying Study on ISO 9001:2000 System
by K. Khoualdi, T.
Alghamdi
Abstract - Database model is an essential topic in software engineering. This paper highlights the technical and commercial differences between object oriented database management system OODBMS and relational database management system RDBMS and follows up some of these differences on a part of ISO 9001:2000 system's database. We apply both models, object oriented database model OODBM and relational database model RDBM, on the same part of ISO 9001:2000 system's database to study how each model represents and accesses the data. This paper provides comprehensive information about OODB and specially guides the students and worker in software engineering to select database model which is matching with their application's requirements.
Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved
Keywords: ISO 9001:2000, Object Oriented Database Model, Relational Database Model.
A
Hybrid Particle Swarm Algorithm for Job Shop Scheduling Problem
by
Gongfa Li, Yuesheng Gu, Hegen Xiong, Jianyi Kong, Siqiang Xu
Abstract - Production scheduling is a hotspot of manufacturing system and the core of the whole advanced manufacturing system. An effective scheduling method and optimization technology is the foundation and the key to realize advanced manufacturing and improve production efficiency. And algorithm research is one of the important content of the production scheduling problem. In recent years, various intelligent computation methods have been gradually introduced into the scheduling problem, such as genetic algorithm and simulated annealing algorithm, etc. The standard particle swarm optimization algorithm has a high computational complexity when used to solve the production of job-shop scheduling problem. The metropolis sampling criteria is introduced into the PSO algorithm. Other algorithms combined with particle swarm optimization algorithm, three kinds of fusion simulated annealing thoughts of hybrid particle swarm algorithm are constructed respectively. Comparing the results of hybrid PSO with the other algorithms in scheduling the job-shop benchmarking problem, the effectiveness and superiority of the hybrid Particle Swarm algorithm are verified.
Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved
Keywords: Job Shop Scheduling, Particle Swarm Algorithm, Simulated Annealing Algorithm.
Customer
Classification Based on Data Mining
by Peishuai
Chen, Chonghuan Xu, Fuguang Bao
Abstract - With the increasingly fierce market competition, customer relationship management centered on the customers has been widespread concerned and developed in the research and application. In-depth study of customer and potential customer is the key to maintain the competitiveness in market. Through analyzing uncertainty in the classification of customers based on data mining tools, this paper proposes an improved ID3(Iterative Dichotomiser 3) algorithm based on rough sets, which can be used for customer classification. Experimental results show that the proposed scheme is efficient and has good performance.
Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved
Keywords: Customer classification, rough sets, Attribute reduction, ID3.
Fuzzy
Support Vector Machines Based on Dual Split off Sample Space
by
Hongyan Pan
Abstract – Since traditional support vector Machine is very sensitive to outliers or noises, a fuzzy support vector machine based on dual split off sample space is proposed in this paper. The membership of a sample is determined by the relative relationship among the sample, hypersphere and the convex hull. Firstly, it finds a hypersphere by traditional SVM, and the radius of the hypersphere is determined by the separating hyperplane and class center. Secondly the convex hull algorithm is used to construct the convex hull of the hypersphere. The space of sample is divided into three parts by hypersphere and convex hull. The fuzzy membership is defined according to the relative position of the sample in hypersphere and convex hull. Based the S function, the membership is defined. Compared with the fuzzy support vector machine algorithm based on the relation between sample and cluster center, this method can give relatively accurate membership to a sample which lies in hyper sphere and convex polygon between. It proves that the fuzzy support vector machines based on dual split off sample space is more robust than the traditional support vector machine and the fuzzy support vector machines based on the relation of a sample and its cluster center.
Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved
Keywords: Convex Hull, Dual Split, Hypersphere, S Function.
Product
Line Production Planning Model Based on Genetic Algorithm
by
Guozhang Jiang, Yuesheng Gu, Jianyi Kong, Gongfa Li, Liangxi Xie
Abstract – At
present, iron and steel enterprise develops towards the direction with
many procedure, many process, many variety and many specification, and
reaches hundreds and thousands of product series and product mix, how to
plan, organize and control steel production, production schedule of its
product line are key issues. Mixed production plan model of product line
can be summed up in a kind of network flow plan. According to the
characteristics of network flow plan, the production schedule model of
product network flow is established through describing digraph-connected
graph of production procedure of iron and steel enterprise. Its goal
function is the biggest profit of production of product line, restrain
functions are the capacity limiting conditions, the balanced condition
of the middle peak point, capacity restrain with supply and sell
production and restrain with enterprise procedure process resources.
Several key resource production procedure processes are chosen to
calculate by using standard library function of the Matlab7.0 genetic
algorithm toolbox to program. Penalty function is adopted in the course
of getting solution. These parameters of scale of father population,
crossover probability, mutation probability and penalty factor are
combined and optimized. Results indicate that goal value reach
convergence after finish 119 iterative operations. It accords with the
actual conditions of this enterprise basically that the optimization
solution to production plans of real iron and steel enterprise by using
the algorithm.
Copyright © 2011 Praise Worthy Prize S.r.l. -
All rights reserved
Keywords: Production Planning Model, Product Line, Genetic Algorithm, Iron and Steel Enterprise, Simulation.
Moderating
Effect Analysis of Relationship Length in Relationship Benefits Using
Data Mining
by
Qingmin Kong, Xiuqing Liang
Abstract - In recent
research and literature, as the relationship marketing becomes an
important part of enterprises’ marketing activities, there come the
concept of relationship benefits. However, it is still not clear that
whether relationship length have effect in the impact of relationship
benefits on relationship outcome. Relationship length has significant
correlation with relationship benefits and relationship outcome,
different relationship length customers have different benefits
perceive. This paper uses data mining to analyze the moderating effect
of relationship length in relationship management. This paper firstly
uses factor analysis to explore the basic dimensions of relationship
benefits, and then use regression analysis to analyze the relationship
benefits preference and the moderating effect of different relationship
length customers. After this, this paper uses structural equation
modeling to analyze the different impact of relationship benefits on
relationship outcome among different relationship length customers.
Finally, this paper proposes a relationship benefits management
strategy. The findings of this paper are directions for future research
and managerial implications.
Copyright © 2011 Praise Worthy Prize S.r.l. -
All rights reserved
Keywords: Relationship Benefits, Relationship Length, Relationship Outcome, Data Mining, Factor Analysis, Regression Analysis, Structural Equation Modeling.
Human
Action Recognition Using Local Features in Bag of Keypoints Paradigm
by
Reyhaneh Rashidi, Javad Haddadnia
Abstract - Recently
local-based approaches have shown promise for computer vision tasks and
local space-time features have become a popular video representation for
human action recognition. In this paper, we focus on recognizing the
action of a person based on the appearance and motion information by
constructing spatio-temporal features in the “bag of keypoints”
paradigm. Our method is motivated by the recent success of bag of
keypoints representations for object recognition problems in computer
vision. We evaluate performance of some local appearance detectors and
descriptors along with the Histogram of Oriented Optical Flow as a
motion descriptor. Here bag of keypoints histograms are computed based
on three diverse descriptors for videos. These features consist of SIFT,
SURF and Harris-PHOG, each of which is concatenated with the HOOF of
interest points. We use Support Vector Machine as the classifier and
evaluate results using two kinds of kernels, the polynomial kernel and
the radial basis function kernel. It is demonstrated that the proposed
methods achieve higher or comparable accuracies as compared with several
state-of-the-art categorization methods on challenging benchmark dataset
of KTH.
Copyright © 2011 Praise Worthy Prize S.r.l. -
All rights reserved
Keywords: Human Action Recognition, Bag of Keypoints Model, Local Features, Optical Flow.
Virtual
Hand Modeling and Simulation Based on Unity 3D
by Lam Meng Chun,
Haslina Arshad
Abstract - In virtual environments, virtual hand is an important avatar for user in direct user interaction technique. Direct user interaction provided an intuitive and natural interaction between user and computer, it is an interaction technique widely used in virtual reality application such as virtual assembly, virtual training system, sketching, virtual object manipulation and so on. This paper presents the modeling of a virtual hand model consists of three layer model which are hand shape, skeleton and skin. Virtual hand is built upon the anatomic structure of the human hand and enhanced visual realism by applying texture mapping. The whole modeling process is done by using 3ds Max software. We also present virtual hand simulation controlled by 5DT Data Glove in Unity 3D. Unity 3D is a game development software and it is also suitable for virtual reality application development. The virtual hand is imported in Unity3D and is programmed with C# language in Visual Studio. The current gesture performed by user can be determined by using 5DT Data Glove SDK, data is collected from the data glove and grasping and releasing simulation is performed in Unity3D. Finally, this paper also demonstrates the data collection process from flock of bird tracking device.
Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved
Keywords: Virtual Grasping, Virtual Hand, Unity 3D.
FDTD
Simulation Method Using Double Signals for Electromagnetic Field: Determination
of the Tissue Biological Characteristics
by Mondher Chaoui,
Moez Ketata, Mongi Lahiani, Hamadi Ghariani
Abstract - This paper describes the method to determine the frequency of an UWB (Ultra Wide Band) wave at which the wave can cross the biological tissue with limited losses and deformation. A study on the influence of frequency variation in signal attenuation is carried especially at interfaces air-tissue and tissue-air. The biological tissue which is characterized by its thickness, its relative dielectric constant and its conductivity, is a sample of fat, localized at a distance of the antenna source which guarantees the inexistence of the attenuation at the amplitude of the electromagnetic fields in the air region. The thickness is about 2.8 cm but the conductivity and the dielectric constant vary with the frequency. The sample is localized in the middle of interest area and it is 50 cm away from transceiver antenna which send signal. A comparison is carried while applying double signals, one is a Gaussian monocycle pulse and other is a simple Gaussian pulse plane and uniform which normally tackles the surface of the samples. To calculate the attenuation of the wave in the biological tissue, we must simulate their propagation by solving Maxwell's equation in these environments. Several methods have been used such as Finite Element Method (FEM) and the Finite Differences Time Domain (FDTD) method.
Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved
Keywords: Attenuation, Electromagnetic Wave Propagation, Finite Differences Time Domain (FDTD), Human Body Tissue, Ultra Wide Band (UWB) Wave.
Forecasting
Job Cycle Time in a Wafer Fabrication Factory by the FPCA-FBPN Approach
by Toly Chen
Abstract - Principal component analysis (PCA) is a multivariate statistical analysis method. This method constructs a series of linear combinations of the original variables to form a new variable, so that these new variables are unrelated to each other as much as possible to reflect information in a better way. A fuzzy PCA and fuzzy back propagation network (FPCA-FBPN) approach is proposed in this study for forecasting the cycle time of a job in a wafer fabrication factory, which is a critical task to the wafer fabrication factory. For evaluating the effectiveness of the proposed methodology, production simulation is also applied in this study to generate some test data.
Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved
Keywords: Wafer Fabrication Factory, Fuzzy Principal Component Analysis, Back Propagation Network, Cycle Time.
A
LED Landscape Lamp Automatic Real-time Control Schema Based on Fuzzy Neural
Network
by
Daiyun Weng, Li Yang
Abstract – Energy-saving, convenience and economy are most important issues of sustainable development of urban landscape lamp in the future. To address on currently popular LED landscape lamp, a kind of automatic real-time control system of LED based on fuzzy neural network was designed, including lightness level conditioning based on fuzzy neural network and lighting effect adjustment based on hybrid strategy. As to lightness level conditioning, multi-layer forward BP neural network control model was built and feedback as well as parameter adjustment methods were also provided. For lighting effect adjustment, hybrid strategy of PWM-based step-less dimming and group control based on serial bus was adapted to implement LED landscape lamp step-less dimming and group control. The LED landscape lamp automatic control schema achieve better environment protection benefit while lighten the city.
Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved
Keywords: LED, Landscape Lamp, Fuzzy Neural Network, Serial Group Control.
Hardware
Designs and Architectures for Projective Montgomery ECC over GF (p) Benefiting
from Mapping Elliptic Curve Computations to Different Degrees of Parallelism
by
Mohammad Al-khatib, Azmi Jaafar, Zuriati Zukarnain, Mohammad Rushdan
Abstract – This paper proposes several hardware design schemas for Montgomery ECC over GF (p). Three known projective coordinates were used to apply ECC computations, in order to eliminate the long time inversion operation, which are Homogeneous, López-Dahab, and Jacobean coordinates. Unlike the usual serial design implementations, our proposed designs utilize the inherent parallelism in ECC computations by using parallel hardware design, in which the ECC computations are performed in parallel to improve the performance of ECC. In addition to the performance, other factors affecting the design of ECC such as area, resources-consuming, system utilization, AT2, and AT factors were studied and enhanced. This was achieved via proposing different design schemas for Montgomery ECC by varying the degree of parallelism for ECC computations. Our proposed designs provide attractive trade-off between mentioned factors, which helps in selecting the most efficient design choices for different security applications based on the requirements and availability of resources for particular application. The performance level of homogeneous coordinates when applied using 4-PM design was five multiplication cycles, which overcomes the other coordinates systems, as well as the known serial design. The 2-PM design for homogeneous projection, on the other hand, improves the system utilization considerably with less area and resources in comparison to the designs with higher degree of parallelism.
Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved
Keywords: Elliptic Curve Crypto-system, Point Doubling, Trade-off, Resources, Area, Hardware Design.
A
New Digital Technique in Predicting Implant Size for Pre-operative Planning in
Total Hip Arthroplasty
by Azrulhizam
Shapi’i, Riza Sulaiman
Abstract -
Digital templating in total
hip arthroplasty (THA) is the preparation process of the pre-surgery
scenario by using digital implant and x-ray. According to surgeons from
Medical Center of Universiti Kebangsaan Malaysia (UKMMC), conventional
methods are still used to find a suitable implant for the patient.
Therefore, a digital method should be developed so that the implant size
detection process can be effectively implemented. This paper will show
how the implant was designed for use in a digital environment. Manual
implant templates used by UKMMC are used as the basis for the implant
design. To match the digital implant and X-ray images with the computer
screen, the pixel ratio between computer display and the digital X-ray
images of patients is generated using the resolution information and
X-ray image pixel density. The pixel ratio is then used in magnification
algorithm to enlarge or reduce the size of a digital implant. A total of
20 X-ray patients were randomly selected to test the accuracy and
effectiveness of the developed digital technique. Results showed this
technique predicted stem component size well, with 55% within similar
size and acetabular component was predicted slightly better with 60%. In
addition, by using digital implant with high resolution, X-ray image
enlargement can be up to 100% successful without affecting the image of
the digital implant.
Copyright © 2011 Praise
Worthy Prize S.r.l. - All rights reserved
Keywords: Digital, Total Hip Arthroplasty, Implant, Size, Scaling, Magnification.
Improvement
of Wood Ultrasonic CT Images by Using Time of Flight Data Normalization
by
Honghui Fan, Hongjin Zhu, Guangping Zhu, Xiaojie Liu
Abstract - The
maximum likelihood expectation maximization (ML-EM) algorithm was
applied to ultrasonic time of flight (TOF) computed tomography (CT) for
wooden pillars. The sound velocity changes with the direction of the
ultrasonic propagation path, therefore, when the image was reconstructed
by TOF data based on ML-EM method, the image had many artifacts. For the
purpose of reducing the artifacts, we proposed a "gap average velocity"
method in imaging process. Transmission paths of an ultrasonic wave
through a cross-section of wood were corrected after TOF data
normalization. The feasibility of TOF data normalization and TOF data
interpolation were examined in detail by using wooden phantoms. The
artifacts evidently disappeared and high quality reconstructed images of
wooded pillars were improved by the proposed technique.
Copyright © 2011 Praise Worthy Prize S.r.l. -
All rights reserved
Keywords: Ultrasonic computed tomography, Maximum likelihood expectation maximization, Time of flight, Normalization.
International Review on Computers and Software - Papers- Part B
Outline-based
Text Image
by
Guorong Xiao, Xuemiao Xu
Abstract - Text
image is a graphic design technique that consists of pictures pieced
together from the printable characters. It is very useful in the
currently text-based communication channels. Text image can be roughly
divided into two major styles, tone-based and outline-based. Some
programs allow one to automatically convert an image to text characters.
However, they can only generate the tone-based text image as the
tone-based one can be regarded as a simple dithering process. This paper
propose a novel method to generate outline-based text image with the
minimal number of characters, given the reference image and focus on the
fixed-width characters used in traditional text image and ignore the
proportional fonts. Convincing results are shown to demonstrate the
effectiveness of the proposed method.
Copyright © 2011 Praise Worthy Prize S.r.l. -
All rights reserved
Keywords: Text Image, Shape Similarity, Grid Deformation.
Integrated
Price Forecast Based on Dichotomy Backfilling and Disturbance Factor Algorithm
by
Quanyin Zhu, Suqun Cao, Pei Zhou, Yunyang Yan, Hong Zhou
Abstract - In order
to improve the accuracy of price forecast on imperfect data by web
extracting, a novel repair data algorithm based on dichotomy backfilling
is proposed in this paper. The price forecast algorithm based on the
disturbance factor is utilized to verify the validity of repair
dichotomy backfilling algorithm. Experiments demonstrated that the mean
absolute errors only can be reduced 1.45 percent. Furthermore, the
repair data algorithm based on average data backfilling and that based
on dichotomy backfilling are explored as well to compare their accuracy.
Experiments demonstrated the reduction of mean absolute errors in the
price forecast verification model. The first one only can reduced 1.56
percent than the actual rate. But the second one can reduced 0.48
percent than the actual rate. Experiment results proved that this
dichotomy backfilling algorithm is meaningful and useful to analyze and
to research the price market on imperfect data by Web extracting.
Copyright © 2011 Praise Worthy Prize S.r.l. -
All rights reserved
Keywords: Dichotomy Backfilling, Imperfect Data, Price Forecast Algorithm, Disturbance Factor.
A
Target Location Method Based on Binocular Stereo Vision for Eggplant Picking
Robot
by
Jian Song
Abstract - A target
location method based on binocular stereo vision is proposed in order to
provide the eggplant picking robot with the three-dimensional
information of the fruit target. The camera imaging geometrical model is
built for the self demarcation of the binocular vision system. The
brightness-based threshold segmentation algorithm is adopted to segment
G-B grayscale images. Such features as the contour profile, area,
centroid, enclosing rectangle, cut-off point of the fruit target are
extracted. The centroid is selected as the matching primitive and The
algorithm with epipolar constraint, uniqueness constraint and disparity
gradient constraint is adopted for the implementation of the fruit
target matching . The eggplant fruit depth information is calculated in
accordance with the homothetic triangle theory. It is measured through
the experiment that the errors of the eggplant depth information mainly
range within ±20 mm in the measurement distance of 300mm~600mm with the
average time used 0.36s. The target location method based on binocular
vision of the eggplant picking robot has simple principle and wide
adaptability, and can be able to meet the requirements for target
location of the picking robot.
Copyright © 2011 Praise Worthy Prize S.r.l. -
All rights reserved
Keywords: Image Processing, Binocular Vision, Target Location, Picking Robot.
Credit
Card Risk Detection Based on Chaos Theory and Cloud Model
by
Yanli Zhu, Yuesheng Gu, Shiyong Li, Shunping Wang
Abstract - Detecting
fraud transactions has been a commonly concerned problem in the
credit-card industry at present, and it directly affects the development
of credit card business. To solve the problems of existing prediction
methods of high false-positive rate, a new detection method of
fraudulent transaction based on chaotic time series has been proposed.
The method, firstly, tries to implement a dynamic customer segmentation
method based on cloud theory to classify the customers properly. Then it
tries to understand the nature of short-term credit card transaction
through time series chaotic analysis and constructs predict model by
employing back propagation neural network, to detect fraudulent
transactions for each kind of clients. The experimental results with
real data demonstrate that the method has lower space complexity and
higher prediction accuracy than the BP neural network prediction.
Copyright © 2011 Praise Worthy Prize S.r.l. -
All rights reserved
Keywords: Risk Detection, Chaos Theory, Phase-Space Reconstruction, Cloud Model, Dynamic Customer Segmentation.
Buyer-Seller
Watermarking Protocol without Trust Third Party
by
Lv Bin, Fei Long
Abstract - Piracy
becomes increasingly rampant in recent years and has caused a great
loss. As a prominent technology, digital watermarking has been developed
to combat piracy. In a real watermark application, the buyer and the
seller follows a secure buyer-seller watermarking protocol, can enable a
seller to successfully identify a traitor from a pirated copy, while
preventing the seller from framing an innocent buyer. A trust third
party (TTP) is required in some known watermarking protocols, which may
affect the security as the seller or the buyer may collude with the
introduced TTP to cheat the other. A buyer-seller watermarking protocol
is proposed in this paper, where no TTP is required without reducing the
security. In the proposed watermarking protocol, homomorphic
cryptosystem and bit commitment are used to achieve the asymmetric
property and solve the disputation. Compared to the previous protocols,
the proposed protocol has good security and can be easily implemented as
no TTP is introduced.
Copyright © 2011 Praise Worthy Prize S.r.l. -
All rights reserved
Keywords: Buyer-Seller Watermarking Protocol, Copyright Protection, Homomorphic Encryption, Bit-Commitment.
Ontology-based
Oracle Bone Inscriptions Machine Translation
by
Jing Xiong, Lei Guo, Yongge Liu, Qingsheng Li
Abstract - Oracle
Bone Inscriptions (OBI) refers to incised ancient Chinese characters
found on oracle bones, which are animal bones or turtle shells used in
divination in Bronze Age China. The vast majority record the pyromaniac
divinations of the royal house of the late Shang dynasty at the capital
of Yin. OBI is the earliest and complete system ancient Chinese
characters. It has very important research merit. With the development
of computer and information technology, digitization processing of OBI
becomes an important aspect in current digitization processing of
ancient Chinese font. But in the digital processing, the burden of OBI
experts is heavy and the knowledge-sharing degree of OBI is low. In
order to resolve these problems, proposed a solution of machine
translation based on ontology. Firstly, the model and processes of OBI
ontology construction are analyzed. Secondly, the flow of ontology-based
machine translation and its key technologies are introduced. Finally,
further research work is prospected.
Copyright © 2011 Praise Worthy Prize S.r.l. -
All rights reserved
Keywords: Oracle Bone Inscriptions, Ontology, Machine Translation, Semantic Similarity, Corpus.
A
Method for Mining Association Rules Based on Cloud Computing
by
Fei Long, Yufeng Zhang, Lv Bin
Abstract - Apriori
algorithm is one of the most classical algorithm to extract association
rules. However, the traditional Apriori algorithm is only suitable for
analyzing and mining the centralized data. Through the analysis of the
Apriori algorithm which had some defects, and the advantages of cloud
computing platform demonstrated in large cluster. To improve the
traditional Apriori algorithm, in this paper, a method for mining
Association Rules Based on Cloud Computing is proposed. The new method
used HDFS to store data and is well adapted to the Hadoop's MapReduce
computing model. It inherits the MapReduce scalability to huge datasets
and to thousands of processing nodes. Experimental results show that it
is very efficient compared with traditional Association Rules algorithm
and have a good speedup when deals with massive data.
Copyright © 2011 Praise Worthy Prize S.r.l. -
All rights reserved
Keywords: Association Rule, Cloud Computing, Apriori Algorithm, Hadoop.
Software
Quality Assurance Based on Java Modeling Language and Database
by
Shukun Liu, Xiaohua Yang, Jifeng Chen
Abstract - The base
and core of software application is software quality. The software
quality assurance is a basic method to solve the problem of software.
During the process of program running, if all the relations in the
program are satisfied, the state of program running is well. The main
skills which based on the dynamic program running trace, the theory of
relational database and technology of stored procedure are showed. These
technologies which can be used to detect the hiding properties among
variants, methods and classes and so on are effective, only when they
are used on the SQL server 2005 platform. The main hiding properties in
the java program are analyzed and the typical methods detecting typical
properties are explained in this paper. The result demonstrates that the
detecting method for testing the close relationships among variables in
java programs is feasible, and these methods can be widely used for
analyzing other programs that can be used in software quality assurance.
Copyright © 2011 Praise Worthy Prize S.r.l. -
All rights reserved
Keywords: Java Modeling Language, Relevance Property, Stored Procedure, Software Quality.
Intrusion
Detection Based on Improved GA-RBF and PCA
by
Yuesheng Gu, Yanli Zhu, Peixin Qu
Abstract - The
intrusion patter identification is a hot topic in this research area.
Using Radial Basis Function (RBF) neural networks to provide intelligent
intrusion recognition has been received a lot of attentions. However,
improper RBF model design may result in a low detection precision. To
overcome these problems, a new intrusion detection approach based on RBF
classifier and improved genetic algorithm (GA) and principal component
analysis (PCA) is proposed in this paper. To alleviate the complexity of
the input vector, the PCA has been employed to eliminate redundant
features of the original intrusion data. In addition, the improved GA
used energy entropy to select individuals to optimize the training
procedure of the RBF. Then, the satisfactory RBF model with proper
structure parameters was attained. The efficiency of the proposed method
was evaluated with the practical data, and the experiment results show
that the proposed approach offers a good intrusion detection rate, and
performs better than the standard GA-RBF method.
Copyright © 2011 Praise Worthy Prize S.r.l. -
All rights reserved
Keywords: Intrusion Detection, Radial Basis Function, Improved Genetic Algorithm, Principal Component Analysis.
Performance
Analysis of Some Known Asymmetric Fingerprinting Schemes
by
Hongyan Wang, Yunyang Yan
Abstract - As an
enhanced version of digital watermarking technology, digital
fingerprinting has been developed to protect digital data from piracy by
embedding a unique serial number called fingerprint into each
distributed copy. To guarantee the fineness of both the merchant and
customer, the fingerprinting schemes are designed to be asymmetric. In
asymmetric fingerprinting, the merchant can track the traitors from a
pirated copy by extracting the embedded unique fingerprint, and the
customer is immune of being framed due to the asymmetric property. A lot
of asymmetric fingerprinting schemes have been proposed in recent years,
and each one has its strengths. To help researchers to further explore
more efficient approaches on asymmetric fingerprinting, the performance
of some known asymmetric fingerprinting schemes is analyzed in this
paper from three aspects, which are computational complexity,
communication cost and the security.
Copyright © 2011 Praise Worthy Prize S.r.l. -
All rights reserved
Keywords: Digital Copyright Protection, Information Hiding, Asymmetric Digital Fingerprinting.
A
Hybrid Price Forecasting Based on Linear Backfilling and Sliding Window
Algorithm
by
Hong Zhou, Quanyin Zhu, Pei Zhou
Abstract - This
paper proposed a hybrid approach for the price forecasting on imperfect
data by web extracting. In this paper the linear backfilling algorithm
is utilized to repair the defect data of commodity prices extracted from
the webpage and the sliding window algorithm is utilized to forecast the
commodity price. Furthermore the linear backfilling algorithm is
compared with the average backfilling algorithm to reveal its validity
in data repair and price forecasting. Experiments demonstrated that
using the linear backfilling algorithm the forecasting accuracy of the
defect data only 0.48 percent inferior than that of the intact data.
Moreover from the experiments it can be seen that the forecasting error
rate produced by the linear backfilling algorithm is 1.08 percent under
that produced by the average backfilling algorithm. In short experiment
results illustrated that this hybrid method is meaningful and practical
to resolve the commodity price forecasting on imperfect data by Web
extracting.
Copyright © 2011 Praise Worthy Prize S.r.l. -
All rights reserved
Keywords: Linear Backfilling Algorithm, Imperfect Data, Sliding Window Algorithm, Price Forecasting.
Implementation
of LDPC Codes Decoding Based on Maximum Average Mutual Information Quantization
by
Lixin Li, Zhengkang Chen, Jie Fan, Nan Qi, Huisheng Zhang
Abstract - As an
enhanced version of digital watermarking technology, digital
fingerprinting has been developed to protect digital data from piracy by
embedding a unique serial number called fingerprint into each
distributed copy. To guarantee the fineness of both the merchant and
customer, the fingerprinting schemes are designed to be asymmetric. In
asymmetric fingerprinting, the merchant can track the traitors from a
pirated copy by extracting the embedded unique fingerprint, and the
customer is immune of being framed due to the asymmetric property. A lot
of asymmetric fingerprinting schemes have been proposed in recent years,
and each one has its strengths. To help researchers to further explore
more efficient approaches on asymmetric fingerprinting, the performance
of some known asymmetric fingerprinting schemes is analyzed in this
paper from three aspects, which are computational complexity,
communication cost and the security.
Copyright © 2011 Praise Worthy Prize S.r.l. -
All rights reserved
Keywords: LDPC Codes, Maximum Average Mutual Information Quantization, Min-sum Algorithm.
Perceptual
Video Content Identification Based on Relative Orientation Invariant between
Geometric Centroid
by
Dayong Wang, Yujie Zhou, Dandan Zhao
Abstract -
Perceptual video content identification is feature vector that uniquely
characterizes one video clip from another. Perceptual means the same
video content should have the same feature vector, and vice versa. In
this paper, a fast and simple video content identification based on
relative orientation invariant between geometric centroid is suggested.
The same perceptual content video should have the same centroid,
further, the orientation of the centroid is identical, and the relative
orientation keeps unchangeable. This method saves time, because it
doesn’t directly convert the original video into a new one with a fixed
frame rate and doesn’t resize the video, but calculates the orientation
and converts the original orientation into a new orientation with a
fixed frame rate. The experimental results show that the proposed video
content identification method is robust, practical and has a good
discriminatory.
Copyright © 2011 Praise Worthy Prize S.r.l. -
All rights reserved
Keywords: Perceptual Video Content Identification, Feature Exaction, Relative Orientation.
Optimal
Contention Window Adjustment for Asymmetry Traffic over IEEE802.11 WLANs
by
Zhengyong Feng, Guangjun Wen, Yuesheng Gu
Abstract - The
IEEE802.11 Wireless Local Area Networks (WLANs) are becoming more and
more pervasive due to the easy channel access mechanism Carrier Sense
Multiple Access with Collision Avoidance (CSMA/CA), but this mechanism
results all nodes including Access Point and other Stations have the
same channel access probability. This characteristic is not suitable to
infrastructure mode which has so many downlink flows to be transmitted
at the Access Point, as a result there is more likely congestion
occurring at the Access Point. To resolve this asymmetry traffic
problem, we develop an Optimal Contention Window Adjustment method. This
method is easy implementation and compatible with the original CSMA/CA
method. It holds the ratio of downlink and uplink flows and at the same
time achieves the maximum saturation throughput in the WLANs. We use the
Markov Chain analytical model to analyze its efficiency and validate it
through simulations.
Copyright © 2011 Praise Worthy Prize S.r.l. -
All rights reserved
Keywords: Markov Chain, Asymmetry traffic, Saturation throughput, WLANs, CSMA/CA.
Sound
Analysis for Diagnosis of Children Health Based on MFCCE and GMM
by
Chunying Fang, Haifeng Li, Wei Zhang, Bo Yu
Abstract - Sound
diagnosis analysis is now attracting more and more researchers in the
world. The purpose of this paper is to study the new method of sound
diagnosis to distinguish and characterize abnormal auscultation.
firstly, this paper proposes a new feature parameter MFCCE based on MFCC,
secondly, finds a appropriate classifier for sound diagnosis through
experiments which compares with the classifier GMM and SVM through
breath recognition, the result is that GMM is better than SVM. Finally,
the further experiments are finished in new recorder data set from
hospital to prove the above ideas correctly. The experimental results
show that the methods have good performance in Nutrition ineffective
cry, cold cough, bronchitis breath, pain cry and non-pain cry
recognition. Thus, this paper presents a method for a quantitative
assessment of children health through sound analysis.
Copyright © 2011 Praise Worthy Prize S.r.l. -
All rights reserved
Keywords: Sound Diagnosis, Mel-Frequency Cepstral Coefficient Entropy (MFCCE), GMM.
Edge
Detection and Noise Reduction for Color Image Based on Multi-scale
by
Feng Xiao, Mingquan Zhou, Guohua Geng
Abstract - The
traditional method of color image edge detection which converts color
image into gray ones and processing the converted gray-scale image edge
does not take into account the color information in color images and
detected result is not so satisfied. The article proposed a multi-scale
edge detection algorithm, which polishing filter for color component
output, extending gradient vector of the polished image edges; selecting
the thresholds of varied multi-scale image edges according to the
improved soft-threshold filtering function and reducing noises and
performing the weighted 2-norm fusion of edges of different-scale-image.
The results show that SNR value is greater than the traditional
algorithm results. The proposed method that takes the color space into
consideration significantly improves the detection effect of
color-image-edge.
Copyright © 2011 Praise Worthy Prize S.r.l. -
All rights reserved
Keywords: Edge Detection, Edge Fusion, Multi-Scale, Wavelet Transform, Noise Reduction.
Robust
Collaborative Tracking in a Multi-camera Surveillance System
by
Weichu Xiao, Weihong Chen, Xiongwei Fei
Abstract - The
distributed architecture of a multi-camera surveillance system is given,
a data fusion scheme using path model and estimated target shape changes
is described, and a collaborative algorithm in distributed multi-camera
surveillance system is proposed in order to track people reliably. Based
on the results of multi-camera data fusion, the proposed algorithm
allocates cameras to objects according to task priorities, the distance
between camera and target and the visibility of targets, which is
characterized by the system assigning cameras to the visible target with
high priority and the nearest distance to the camera. It is called
Priority and Distance Algorithm (PDA). The Quality of Service (QoS)
function is introduced to describe the system performance. Experiment
results show that the proposed method can coordinate cameras to track
people reliably and has good robustness.
Copyright © 2011 Praise Worthy Prize S.r.l. -
All rights reserved
Keywords: Multi-Camera, Collaborative, Tracking, Data Fusion.
Comparison
between Three Algorithms for Smooth Support Vector Regression
by
Bin Ren, Huijie Liu, Lianglun Cheng
Abstract - Smooth
functions can transform the unsmooth support vector regression into
smooth ones, and thus better regression results are generated. It is one
of key problems to seek better smooth function in this field for a long
time. In the paper, using series expansion. The BFGS-Armijo and Newton-Armijo
algorithms have been used to train smooth support vector regression (SSVR),
and the latter has faster speed. Newton-PCG algorithm is just enough
method for unconstrained problem which has better speed than Newton in
theory. On the numerical experimentation of using BFGS-Armijo, Newton-Armijo
and Newton-PCG to train SSVR, this paper gives comparison among the
three algorithms, and obtains that Newton-PCG has the best result.
Copyright © 2011 Praise Worthy Prize S.r.l. -
All rights reserved
Keywords: Support Vector Regression, BFGS-Armijo Algorithm, Newton-Armijo Algorithm, Newton-PCG Algorithm.
PCNN-Histogram
Based Multichannel Image Segmentation Algorithm
by
Beiji Zou, Haoyu Zhou, Geli Lv, Guojiang Xin
Abstract - A new
multichannel image segmentation method based on pulse-coupled neural
network (PCNN) and histogram is presented in this paper. The method
segments image by utilizing PCNN’s specific feature that the fire of one
neuron can capture firing of its adjacent neurons due to their spatial
proximity and intensity similarity. The method judges the weight of each
channel in the multichannel image by a histogram-likely way firstly.
Then it weightedly combines these channels and active the PCNN. During
the iterations of PCNN, the method automatically confirms the best
iteration times by comparing the maximum of variance ratio of each
iteration. At last, the Shannon entropy rule is used to determine the
segmentation results. Experimental results show that the proposed method
performs well in both results and efficiency.
Copyright © 2011 Praise Worthy Prize S.r.l. -
All rights reserved
Keywords: Multichannel Image Segmentation, Histogram, Pulse-Coupled Neural Network.
|
Please send any questions about this web site
to
info@praiseworthyprize.com
|
