Showing papers in "Application Research of Computers in 2005"

Survey on Intrusion Detection Techniques

Abstract: Intrusion detection,as one of the most active and important network security technology,can compensate the shortcomings of traditional security protection measure.Through building dynamic security cycle,it can promote the protection capacity of system and reduce the security threats as great as possible. An overview of basic issues on intrusion detection is shown in this paper,which is involved with three main aspects of intrusion detection,including Intrusion Detection System(IDS) architecture,intrusion detection methods and IDS evaluation.Firstly,three kinds of IDS architectures are analyzed.Then,current intrusion detection models are given,and their merits or shortcomings are discussed in detail.Finally,some future promi￣sing directions are presented.

Image Enhancement Algorithm Based on Retinex Theory

Abstract: Input color image is first divided into three pictures according to its R, G and B values, and each one represents the intensity of reflected light with long-, medium-and short-wavelength. In order to determine the color of each pixel, these three pictures are then processed independently to calculate its lightness. Finally, color in Retinex color space is mapped to normal RGB space to reconstruct enhanced image. Study shows the resulting image often has better color fidelity and larger dynamic range. Computer simulation results demonstrate that it is possible to achieve very good performance by making use of this theory correctly.

A Review of Job Scheduling for Grid Computing

Abstract: This paper addresses the characteristic and performance metric of job scheduling for grid computing, reviews exis￣ting job scheduling technologies and algorithms including heuristic job scheduling, Agent based job scheduling, modeling and analysis using Petri net, cost based job scheduling and load balancing, and finally makes an expectation about its future research directions.

Survey of Automatic Target Recognition and Tracking Method

Abstract: The effective recognition and tracking for extended target in the complex background is a challenged problem. This paper classifies the algorithm of current Automatic Target Recognition (ART). Those features, which have RST invariance and tracking method of moving target, are also discussed. At last, some issues that should be solved in the future are proposed.

Chinese Term Extraction System Based on Mutual Information

Abstract: An automatic Chinese term extraction system based on mutual information is presented.Firstly,the system gets term candidates by calculating internal associative strength of characters string using mutual information. Then, term selection from term candidates is done using basic word dictionary, common collocations suffix, prefix bank and some term POS composing rules. Experiment shows the precision is 72.19%, recall is 77.98% and F measure is 74.97%.

Clustering Ensemble Approaches: An Overview

Abstract: Ensemble approaches are widely and successfully used in classification algorithms and regression modelsIt can offer better results for overcoming instabilities in classification algorithms and regression modelsHowever,in unsupervised learning,the researches of ensemble approaches are concerned only in recent yearsBecause the prior information of data sets in unsupervised learning is unknown,the ensemble approaches of classification algorithms and regression models can't be utilized in the same way directlyRecent researches and experiments show that clustering ensemble approaches can enhance the robustness and stabilities of unsupervised learning greatlyThis paper makes an overview of the clustering ensemble approaches in recent yearsIt illustrates the contents and characteristics of recent clustering ensemble approaches research and discusses the future directions of clustering ensemble study

Survey of Research on Access Control Model

Abstract: Access control is an important information security technology. To enhance benefits and increase competitive power, many modern enterprises have used this technology to secure their information manage systems. In this paper, several main access control models, such as traditional access control models, role-based access control models, task-based access control models, task-role-based access control models, and so on, are discussed and compared in detail. In addition, we introduce a new model called UCON, which may be a promising model for the next generation of access control.

An Energy Efficient Protocol for Wireless Sensor Network

Abstract: A challenge in collecting sensed information is that nodes only have limited battery power.So collecting data in an energy efficient manner is critical to operate the sensor network for a long period of time.This paper analyzes the energy consumption of sensor networks and proposes an energy preserving protocol that reduces the amount of energy spent per cycle.In the protocol,each node communicates only with a close neighbor,delivers messages along a shorter-path and takes turns becoming a cluster head to transmit to the base station,thus prolonging nodes' lifetimes. Simulation results show that it can achieve significant energy savings.

Research of Enterprise Resource Management Supporting Network Manufacturing

Abstract: Lays emphasis on how to realize enterprise resource model for the point above, and build enterprise resource base on the basis of Oracle database. At the same time, illustrates how to set up various management system for the resource workflow to separate the resource data from the workflow and import the new data into resource base, so it ensure integration and management of all kinds of enterprise resource.

Study of Hybrid Genetic Algorithm Based on Simulated Annealing

Abstract: Taking a modified Simulated Annealing algorithm as a genetic operator realized the combination of the local sear￣ching ability of SA and global searching ability of GA. A new hybrid algorithm of Genetic Simulated Annealing had been (designed) with dynamic probability of crossover and mutation, and tested by a nonlinear function optimization. The results indi￣cated the hybrid algorithm can improve significantly the efficiency of GA for solving nonlinear optimization.

Design of Mobile Terminal of Vehicle for Monitoring System Based on GPRS

Abstract: Analyzed the advantage of vehicle real-time monitoring and attempering system by GPRS communication technology,introduced the system components of the vehicle monitoring system and the working principle,in addition,provided how to design the hardware and software of the mobile terminal by MSP430F147 and Sony-Ericsson GR47.

A New Mutual User Authentication Scheme Using Smart Card

Abstract: User authentication is the important problem in network security.Recently,Chien et al proposed an efficient user authentication scheme using smart card,however,it can see that their scheme is vulnerable to a Reflection Attack,Parallel Attack,Replay Attack and so on.Herein,this paper first shows the weakness of Chien et al's scheme,and then proposes a new improved scheme with better security.

Survey of Human Face Recognition

Abstract: Face recognition is an important technique of identification. It will be applied extensively. Characteristics of face recognizing technique in the development of face recognition are briefly reviewed. Pivotal method and technologies are expounded, and merits and demerits of these methods are indicated, according to three main steps of face recognition: face detection and localization, facial feature extraction and face verification and identification. Vista of future research in face recognition technology is distilled.

Survey on Workflow Module

Abstract: Workflow Module, which is considered to be one of the most core part of WFMS (WorkFlow Management System) since CIMS has become very popular in the last few years. The current state of the art of workflow module, including what is a workflow module and its functionality, architectural model, component container, theory and shortcoming etc., are outlined. Lastly, some open issues about workflow are discussed, and the future direction of it is pointed out.

Research of SOA and Its System Building

Abstract: Service-oriented architecture is a new architecture.Analyze the feature and basic structure of SOA detailedly and compare the differences between SOA and object-oriented architecture. With an example, this paper gives the approach architecting a SOA System and describe the process from the component-based design method to the service-oriented design method.

A New Method of Low Contrast Image Enhancement

Abstract: A new method of low contrast image enhancement is presented based on PCNN and nonlinear extension in conjunction with characteristics of human visual consciousness in this paper. Approaches of local and whole contrast enhancement of image are discussed in detail. It is showed that the method not only ameliorates visual effect obviously but also has properties of smoothing image, enhancing image and self-adaptation through analysis and experiments.

Analysis and Amendment of EAP-AKA Protocol

Abstract: 3G-WLAN interworking is a research hotspot now, and EAP-AKA is its authentication and key agreement protocol. The paper analyses the procedure and security of EAP-AKA protocol, and the vulnerability and possible attack to this protocol are described. To the vulnerability and possible attack, an amended scheme is presented. Finally, an analysis of the amended scheme has been made.

Review on Vehicle Detection and Tracking Techniques Based on Video Processing in Intelligent Transportation Systems

Abstract: Compared with traditional traffic detectors, the video sensor has lots of advantages such as fast response, easy installation and maintenance, the ability to monitor wide areas and obtain more kinds of traffic parameters, and as a result, it has been widely used in Intelligent Traffic System(ITS) in recent years. Up to now, a number of video processing and analyzing methods have been proposed for vehicle detection and tracking. In this paper, vehicle detection and tracking techniques are introduced and categorized based on their key processing techniques (feature, area or model) and processing domain (spatial or temporal), the advantages and disadvantage of each method are also analyzed. Additionally, the problems that still exist and the trend of the research in this field in the future are indicated.

A Fair and Off-line Electronic Payment Scheme Based on Elliptic Curve

Abstract: This paper proposes a off-line electroni c payment scheme based on the elliptic curve cryptosystem. A payment system by Smart card proposed by S.Brands is improved,and an anonymity-revoking electronic payment system is presented.On the one hand, customer’s privacy can be compromised by the bank or by the payee .On the other hand,the anonymity can be removed by a trusted third party with the help of bank.In this case,the third party can link a payment with corresponding withdrawal and prevent money forgery,laundering,badger game and black mailing. The scheme is suitable for application in Smart card because of using elliptic curve cryptosystem.

Study on New Pretreatment Method for Chinese Text Classification System

Abstract: Presents a new text pretreatment method that applying programme flows control to eliminate the single Chinese word, pure English words, number and Chinese words containing English letter or maths symbol from the original text vector Consequently the features that represent the text turn into the pure Chinese term As a result, not only dimension of original text vector is deduced greatly but the information contents of text vector are improved enormously

Trouble Pattern Analyze Based on Data Mining

Abstract: A new trouble pattern analysis was put forward based on thought of data mining.The trouble information dimension table come into being by collecting and cleaning up the fault phenomenon.The association rule dimension table was made up of the technique parameters and the trouble causes.The trouble information are analyzed by frequent item set of the Apriori arithmetic based on the trouble information dimension table and the association rule dimension table.It made sure the causes of the trouble and chose the priority resolving scheme by matching the trouble and building and filtrating the candidates.

Research on Augmented Reality of Tracking Registration

Abstract: The principle of augmented reality of single camera vision based was investigated and the process of marker identify was andlyzed through pattern matching.Based on the discuss of the tracking registration flow frame of augmented reality, virtual training prototype system was developed under OpenGL library after snatching figure from single camera.

Research on Principle and Defense of Buffer Overflow Attacks

Abstract: The buffer overflow attacks have been the most common form in the network attacks and become a predominant problem in the system and network security area. After the principle of the buffer overflow attack is explained, this paper analyzes the causes leading to the attacks. Then the usual attack types the attackers often exploit are presented. Last, the common measures to defend these attacks and my own idea about the solution of this problem by modifying the compiler and the relevant functions of the system are given.

Research of Modeling Method of Administrative Transaction Processing System Based on Role Model

Abstract: Firstly,analyzes the characteristics and the present situation of the administrative transaction processing system briefly. Secondly, referring to some related enterprise models, role model is introduced into the field of government administrative transaction processing system. The paper discusses the role's definition and relationship in the administrative transaction processing system and presents the modeling method of administrative transaction processing system based on the role model. Finally, the contributions and significations of this method are discussed.

A New Meter Dial Indicator Detection Method Based on Center-point Projection

Abstract: The Hough transform is firstly used to detect the center point of the dial. The dail image is then projected to the center point and the dial indicator is finally detected from the image center-point-based projection. Experiments show that the proposed method works very well for the dial indicator detection. Comparing with the classical Hough transform, it has the merit of fast processing speed and very low computation storage.

Overview of Technology of Building Domain Knowledge Based on Ontology

Abstract: With the knowledge resources become more and more abundant, how to organize and manage knowledge efficiently in order to implement the share and reuse of knowledge and make use of knowledge more efficiently become a pressing research task. The construction of knowledge based on Ontology can implement the knowledge share by providing the common understanding of domain knowledge. In this paper, the definition, function, design and application about Ontology are introduced and analyzed. In the end, we discuss the research direction and emphasis in future are discussed.

Research of Video Surveillance System Based on Streaming Media Technology

Abstract: The video surveillance system based on streaming media technology has the advantage of low cost, stable running, unconstrained by firewall, etc.This paper discusses the structure and programming of the video surveillance system based on streaming media, also it analyzes the broadcast delay in detail.

Research on Integration of Multi-source Spatial Data

Abstract: At present, the modes of the integration of multi-source spatial data are data format exchange, data interoperation, and directly data access. In this paper, the author analyzes the principle and the shortage of such modes. Then the seamless integration of multi-source spatial data, universal spatial data engine, and middleware technology are emphasized.

Remote Sensing Image Fusion Based on DCT

Abstract: Applies DCT to the remote sensing image fusion between the panchromatic image and the multi spectral imageFirst,DCT will be performed in the panchromatic image extracting the low freqency from the I component,and the hign frequency from the panchromatic image that the DCT has been performed,fuse both parts to form the fused image in transformation fieldThrough the experiments,definition of the fused image is treated as the exclusive standard judging whether the fused result is optimum

A Method of Commodity Recommendation Based on Customer Shopping Model of Bayesian Network

Abstract: Presents a new recommendation framework based on customer shopping model. This framework formalizes the re￣commending process as knowledge representation of the customer shopping information and uncertainty knowledge inference process. Firstly,this approach builds a customer model of Bayesian network by learning from customer shopping history data, then presents a recommendation algorithm based on probability inference in combination with customer present shopping action. Experimental results demonstrate that this method can effectively and in real time generate an individual recommendation set of commodity, it is better than some traditional methods in rates of coverage and precision.