Table of Content

01 October 2007, Volume 27 Issue 10

Previous Issue    Next Issue

Database Technology

Research of main memory database

2007, 27(10):  2353-2357. 

Asbtract ( )   PDF (1062KB) ( )  

Related Articles | Metrics

With the price of random access memory dropping and the 64 bit CPU being used widely, the techniques about Main Memory DataBase (MMDB) are studied prevailingly. And a few of prototype systems and business systems have been developed. This paper analysed the difference between the MMDB and the disk resident database in terms of query processing and index techniques, etc. This paper summarized the current research about main memory database and the open problems of MMDB. At last, it points out some directions for future research.

Discover formulas based on GEP with trans-gene

2007, 27(10):  2358-2360. 

Asbtract ( )   PDF (704KB) ( )  

Related Articles | Metrics

When mining knowledge by traditional Gene Expression Programming (GEP), users hardly control the process, quality and speed of evolution. To solve this problem, the authors borrow the idea of Transgene from modern biology, import it into formulas mining based on GEP, and get series of results. This article surveys the achievements in Gene Expression Programming with transgene, including gene injection techniques, process for trans-gene, evolution intervene, etc. By fusing natural and manual selection, during evolution process, user can guide the evolution towards expected direction.

Survey of research on anonymilization technology in data publication

2007, 27(10):  2361-2364. 

Asbtract ( )   PDF (859KB) ( )  

Related Articles | Metrics

Anonymilization, as known to all, is a good method to solve the leakage of confidential information, once after it was proposed. The research on anonymilization technology can be divided into two categories, one on the generalization principle, and the other on the implementation ways. In this paper, we gave an overview study of the status quo and trend of the research, and discussed the open and challenging problems in anonymilization technology.

Web Service retrieval based on operation clustering

2007, 27(10):  2365-2368. 

Asbtract ( )   PDF (825KB) ( )  

Related Articles | Metrics

By taking full advantage of the service description information, an approach for modeling Web services at the operation level was proposed. Base on the definition of service co-relevance and its measuring approach, a variant k-means clustering method has been developed to group service operations into different operation concepts. A service retrieving model and the corresponding algorithms have also been devised using the operation concepts. Experimental results show that the proposed approach greatly improves the performance of retrieving Web services.

Algorithm of outliers mining based on grid techniques in high dimension large dataset

2007, 27(10):  2369-2371. 

Asbtract ( )   PDF (772KB) ( )  

Related Articles | Metrics

An algorithm of outliers mining based on grid techniques in high dimension large dataset called Outliers Mining Algorithm based on Grid Techniques （OMAGT） was proposed. Focusing on the distributing characteristics of high dimension large dataset, clustering regions were found out by using the way based on grid techniques, moreover, those clustering dataset unable to turn into outliers in clustering region were deleted. Then outliers mining was done using algorithm Local Outlier Factor （LOF） in the remaining datasets. In the algorithm OMAGT, dynamical release of clustering information was preferably carried out. Thus, information of reserved outliers mining was restricted in limited memory capacitance, so both time efficiency and space efficiency were improved. Results in both theory analyses and experiments show that this algorithm is feasible and efficient.

Multi-resolution online classification algorithm for data streams

2007, 27(10):  2372-2375. 

Asbtract ( )   PDF (829KB) ( )  

Related Articles | Metrics

An incremental classification algorithm based on nearest neighbor technology, which adapts to the sudden concept shift over data streams, was proposed. The algorithms uses grid technique to quantize the feature space of data set and uses a multi-resolution data representation based on Haar wavelets to find adaptive class label of a test point. Experiments performed on both synthetic and real-life data indicate that the proposed classifier outperforms existing algorithms for data streams in terms of accuracy. The algorithm's low update and computational cost makes it highly suitable for data stream applications.

Classification rule acquisition based on extended concept lattice

2007, 27(10):  2376-2378. 

Asbtract ( )   PDF (622KB) ( )  

Related Articles | Metrics

Concept lattice is a powerful tool for data mining and rule acquisition. Through analyzing the characteristics of concept in concept lattice, extended concept lattice and classification rule acquisition based on extended concept lattice was proposed. Experimental results show that this algorithm could obtain simple and understandable rule set.

Research of log overflow protection based on bitmap

ran Zheng ZhanHuai Li YanLong Wang

2007, 27(10):  2379-2382. 

Asbtract ( )   PDF (794KB) ( )  

Related Articles | Metrics

On the basis of the database system and data disaster tolerance system, a bitmap-based log overflow protection mechanism was designed due to the problem of the size limitation in the current log technology. We discussed the implementation of the mechanism from data consistent and atomic operation, and analyzed the superiority of the mechanism compared with traditional log technology. Its proved by the prototype that the mechanism effectively resolves the log overflow problem due to network congestion and sudden increment of I/O requests, thus data replication can be guaranteed.

Implementation of adaptive aggregate operator based on synopsis

HaiSheng TIAN

2007, 27(10):  2383-2387. 

Asbtract ( )   PDF (972KB) ( )  

Related Articles | Metrics

In this article we proposed a Batching Processing algorithm based synopsis, which can adjust its processing policy according to the arriving data rates in order to achieve better query quality under limited system resource. In addition, according to the characteristics of aggregate we can use some optimized algorithm to gain real time or nearly real time query results. We present prototype-based experimental results on the Argus DSMS to characterize the efficiency and effectiveness of our approach at the end of this paper.

Network and communications

Analysis of AODV in MANET-based NCS

Chun-Xue WU Wen-Dong HAN

2007, 27(10):  2388-2390. 

Asbtract ( )   PDF (596KB) ( )  

Related Articles | Metrics

Based on the analysis of route discovery and route maintenance algorithm in Ad hoc On-demand Distance Vector （AODV）, combined with the data transmission characteristic of MANET-based Networked Control Systems （NCS）, we modified the route maintenance algorithm to reduce the average end to end delay, by means of starting route discovery immediately by the influenced source node on detecting the link failure. Simulation shows the effectiveness of the modified protocol when the MANET-based NCS is in small scale, and it is helpful to improve the system stability.

Prediction-based resource matching algorithm

Zhi-Gang HU Zhou-Jun HU

2007, 27(10):  2391-2394. 

Asbtract ( )   PDF (840KB) ( )  

Related Articles | Metrics

Resource matching, a procedure of Grid task scheduling, aims at finding appropriate resources for Grid tasks according to its requirements. The resource information recorded in Grid Resource Information Service （GRIS） is always static. Because of dynamic characteristic of local jobs arrived in a Grid site, that some of the resources in candidate resource set determined based on static information can not satisfy tasks QoS requirements. This paper predicted resource future status based on its historical dynamic information. Grid task average completion time and distribution of completion time were given. According to tasks QoS requirements and considering both current and future status of a resource, a resource matching model and a matching algorithm were presented. The experimental results show that the algorithm can reduce amount of candidate resource effectively, which can decrease time complexity of a scheduling.

Routing selection algorithm under multiple link state limited

Weiya Wang

2007, 27(10):  2395-2397. 

Asbtract ( )   PDF (560KB) ( )  

Related Articles | Metrics

The Combination of Genetic Algorithm and Ant Colony Algorithm inherits the advantages of Genetic Algorithm and Ant Colony Algorithm, and has a higher efficiency than Genetic Algorithm and a faster speed than Ant Colony Algorithm in computing the shortest route under multi link state limited. The testing proves that the Combination Algorithm is a new better heuristic algorithm both in the efficiency of precise results and time.

Performance evaluation of real-time video transport over UAV multi-hop Ad Hoc networks

Si-Wei LIU Yu-Ting ZHAO

2007, 27(10):  2398-2400. 

Asbtract ( )   PDF (577KB) ( )  

Related Articles | Metrics

The performance of real-time video transport over Unmanned Aviation Vehicle (UAV) multi-hop Ad Hoc networks (UMAN) was evaluated on a novel hardware in the loop simulation platform. The test video was encoded in the Motion JPEG or the MPEG-2 format with the frame rate of 25fps. Tests show that the MPEG-2 codec can provide real-time video service within 3 hops, while the Motion JPEG codec cannot support multi-hop real-time video transport. The packet delivery ratio (PDR) and the decoded block ratio (DBR) got with the MPEG-2 codec has an average increase of 55.688% and 205.643%, compared with the results got with the Motion JPEG codec, respectively. The UMAN can extend the operational range of the UAV system to 4~9 times of the range offered by one UAV with the capability of supporting real-time video transport at the same time.

Application of relaxation in TCP friendly rate control

Da-min ZHANG Lin-zhou-ting CHEN

2007, 27(10):  2401-2402. 

Asbtract ( )   PDF (591KB) ( )  

Related Articles | Metrics

As TCP Friendly Rate Control (TFRC) is used in the application of the Real-Time Streaming Media, the fluctuation of the sending rate is obvious. It is harmful to the transmission of the Real-Time Streaming Media. This essay employed Relaxation Algorithm to limit the variation of the sending rate of the TFRC flow adaptively. It made the sending rate of the TFRC flow more smooth and constringent. The experimental results show that Relaxation Algorithm can improve the capability of TFRC while guaranteeing the high quality of transmission of the Real-Time Streaming Media.

Semi-global trust model design for Peer-to-Peer networks

Jie ZHANG Zheng Zhao Xiao XIONG

2007, 27(10):  2403-2405. 

Asbtract ( )   PDF (580KB) ( )  

Related Articles | Metrics

The novel trust model for P2P file sharing systems wasintroduced. First, an objective criterion to compute each peers reputation was gaven. Second, based on their values we divided all the peers into 2 parts: one is super peers who share more resources and service and also been given some privileges; the other is normal peers who act normal or new comers, they construct an unstructured network. Third, a semi-global structure for computing and storing the trust value was described. Finally, the experiments show that the feasibility and robustness of this reputation management system.

Research of a deterministic small world network based on Cayley graph

Zhen-Tao SUN

2007, 27(10):  2406-2408. 

Asbtract ( )   PDF (556KB) ( )  

Related Articles | Metrics

A deterministic small-world network based on Cayley graph which shows local clustering and low network diameter was proposed. And then analyze some main properties were analyzed and a routing algorithm was developed. At last, validity of this network was verified by experiment.

On-demand multicast routing protocol with on-demand clustering in Ad Hoc network

2007, 27(10):  2409-2411. 

Asbtract ( )   PDF (597KB) ( )  

Related Articles | Metrics

ODMRP protocol is an on-demand multicast routing protocol in Ad hoc network. This paper presented the On-Demand Multicast Routing Protocol with On-Demand Clustering. ODC-ODMRP introduces on-demand technique to reduce the flooding control overhead and improves the performance of networks. At the same time, ODC-ODMRP reserves other key merits of ODMRP such as high performance in conditions with frequently changing topology. Analysis and simulation show that the protocol is characterized with its quick convergence and low maintenance costs of routing, robust and efficient flooding platform over dense, large mobile networks.

Improved SNMP-based network topology discovery algorithm and its implementation

an-yi yang

2007, 27(10):  2412-2413. 

Asbtract ( )   PDF (604KB) ( )  

Related Articles | Metrics

An improved SNMP-Based network topology discovery algorithm was put forward. This algorithm could, based on MIB-Ⅱand TCP/IP protocol addressing principles, accelerate the speed of topology discovery and reduce the network loads. This system was developed at CERNET2. The results of experiments were presented on a map. The results show that this system can discover the network topology quickly and exactly.

Application of SVM based NFD in network management platform

2007, 27(10):  2414-2416. 

Asbtract ( )   PDF (544KB) ( )  

Related Articles | Metrics

To overcome the deficiency of existing Networks Fault Diagnosis (NFD) scheme in real-time, forecast and intelligence. Networks Fault Diagnosis scheme was designed to apply to the Network Management Platform. The scheme is an application of SVM multi-classification based on binary tree. The efficiency of network monitoring and fault management was improved. The feasibility and availability of this scheme were proved through the test. Powerful capability of early warning was presented to the small samples in the actual measurement; meanwhile higher classifying precision is also to the multiclass NFD.

Improved algorithm for streaming media congestion control based on feedback

2007, 27(10):  2417-2419. 

Asbtract ( )   PDF (570KB) ( )  

Related Articles | Metrics

The streaming media congestion control algorithms, which employ sending rate control, can not efficiently solve feedback delay and parameters settings. Based on the Feedback Congestion Algorithm (FCA), this paper presented an improved algorithm (IFCA) on feedback real-time performance and parameter adjusting. Simulation results show that the proposed method, by which jitter and throughput are improved, can not only be more suitable for streaming media, but also keep TCP-friendly.

Evaluating method over remote Web based on linear weighted technique

2007, 27(10):  2420-2422. 

Asbtract ( )   PDF (634KB) ( )  

Related Articles | Metrics

To the questions of evaluating method over remote web, a technique for evaluating method over remote web based on Linear Weighted Technique (LWT) was proposed, which was implemented and achieved with VBScript programming language and dynamic web technique. In succession, the build of the architecture, the design of functional modules of query, sort, calculation and user administrators, the framework of background database were made. As to the kernel issue of evaluating method over remote web, i.e. how to implement and achieve LWT with VBScript programming language in the environment of Active Server Pages routine was described in detail. The feasibility and validity were checked in a case study. The big advantage of the technique is that it can synthetically consider the effects of various evaluating factors on evaluating results with the better objectivity and equity.

Power-aware routing mechanism based on Newton's interpolation in MANET

2007, 27(10):  2423-2425. 

Asbtract ( )   PDF (810KB) ( )  

Related Articles | Metrics

One main constraint in Mobile Ad Hoc Network (MANETs) is limited by the power of the node. Therefore, much effort has been paid to reduce power consumption. A new power-aware routing mechanism is presented based on Newton's interpolation. According to the remaining energy and the traffic load at nodes, the life of the nodes was calculated out, and then from the bearable life of nodes, one minimum power drain routing can be choused. In this way, the mechanism not only takes account of the balance of the power at nodes, but also pays attention to the minimum power drain routing problem of the whole network. Simulations show that, compared to previous algorithms, it has better performance.

Secure clustering algorithm based on trust evaluation in Ad Hoc network

Jing Zhang li Xu daowei Xu

2007, 27(10):  2426-2429. 

Asbtract ( )   PDF (814KB) ( )  

Related Articles | Metrics

Ad Hoc network has particular characteristics. In this paper, a new trust evaluation scheme based on clustering was proposed. Each node chooses one node that has the highest value as its cluster head. It is proven in theory and in experiment that this algorithm is more secure then other schemes for clustering based Ad Hoc networks.

Grid resource recovery based on ant algorithm

2007, 27(10):  2430-2432. 

Asbtract ( )   PDF (623KB) ( )  

Related Articles | Metrics

Grid resource recovery research based on ant algorithm was proposed, which used ant algorithm as strategy to achieve a suboptimal solution in grid resource recovery. In this paper, the initialization and updates of ant algorithm Pheromone were improved as well as its implement grid resource recovery and resolve grid resource's characteristics commendably, and then it can be completely verified with Matlab.

Mixed concurrency control algorithm of composed service transaction based on conflict probability

huiqiong zeng

2007, 27(10):  2433-2436. 

Asbtract ( )   PDF (820KB) ( )  

Related Articles | Metrics

Because Web Services are relaxed coupled and independent to each other, two single services in a composed service can execute concurrently. In allusion to the characteristic, combining the advantages of optimistic and pessimistic concurrency controls, this paper presented a mixed concurrency control mechanism on occurrence probability of conflict, which can improve the degree of concurrency. Taking full account of concurrency control of instances which are form the same composed Web Services or different composed Web Services, the mechanism can ensure correctness and consistency when multi instances of composed web service execute concurrently.

New TCP algorithm over heterogeneous networks

Da-peng QU

2007, 27(10):  2437-2439. 

Asbtract ( )   PDF (720KB) ( )  

Related Articles | Metrics

A new TCP algorithm that was applicable to heterogeneous networks was proposed. It can not only find the best point in performance and keep the network load around it, but also determine the reasons of dropping packets and adjust parameters reasonably to avoid the mechanism of blind halving congestion window in TCP-Reno. The experimental results show its performance is better than TCP-Reno and other improved protocols of the same kind.

Information security

Research of intrusion detection based on clustering and support vector machine

2007, 27(10):  2440-2442. 

Asbtract ( )   PDF (659KB) ( )  

Related Articles | Metrics

A new approach based on clustering and Support Vector Machine （SVM） was proposed. It could reduce the computational complexity and get better performance in the intrusion detection. Firstly, clustering was imposed on data set using K-MEANS partition algorithm and clusters were established. Secondly, the support vector was searched only when the labels in a cluster were different. Finally, the experimental results using KDD CUP 1999 demonstrate the efficiency of our approach.

Intrusion detection system based on PC-LINMAP and cloud theory

Qiu-yu ZHANG Lei SUN

2007, 27(10):  2443-2445. 

Asbtract ( )   PDF (558KB) ( )  

Related Articles | Metrics

A new decision method to judge the damage degree of system intrusion possibility was proposed, which was based on PC-LINMAP and cloud theory. Firstly, the index of system performance will be given weighted values by using PC-LINMAP Theory, and these computed weighted values will be compared and fused with the values out from index of system performance under ideal conditions. Then this result of comprehensive assessment under ideal conditions could be obtained. Following next, the variation could be obtained by the comparison of real time's weighted fusion values with comprehensive assessment under ideal conditions at each specific moment. Finally, the Cloud Generator which is based on Cloud Theory combined with the above variation can achieve the qualitative description about the damage degree of current system intrusion possibility. The result of experiment shows that this method is effective.

Alert aggregation algorithm based on category and similarity

2007, 27(10):  2446-2449. 

Asbtract ( )   PDF (857KB) ( )  

Related Articles | Metrics

The article proposed an alert aggregation algorithm, which categorized all alerts into four classes on the basis of attack event type, and each attribute belonged to one of the four buckets according to the characteristics of the value of the attribute. Different methods were used to calculate the similarity of the values of the attributes and different threshold values were configured in different situations. An external interface was also implemented to tune the threshold dynamically at run time. The percentage of each attribute in the final result depended on the category of the alert. Aggregation or not depends on the comparison result between the maximum similarity and the pre-defined threshold.

Multi-secret key agreement scheme with threshold and authority

Yanshuo Zhang Zhuojun Liu

2007, 27(10):  2450-2452. 

Asbtract ( )   PDF (574KB) ( )  

Related Articles | Metrics

Key management is an important field in information security. Key agreement is one of the core problems in the key management．A new computationally secure threshold multi-secret key agreement scheme with authority based on the theory of linear codes was introduced．The scheme makes the users share the secret key, and does not need to be based on discrete logarithm. The set of participants to perform key agreement should satisfy threshold. In the scheme, multi-secret key can be initiated in one key agreement session. The scheme based on linear codes has the property of authority by checking the equality to prevent the third party attack.

Adaptive image watermarking method based on chaotic encryption

2007, 27(10):  2453-2455. 

Asbtract ( )   PDF (612KB) ( )  

Related Articles | Metrics

According to the charactersitics of the digital watermark, this paper proposed a kind of wavelet domain image digital watermarking technique using two-dimensional chaotic encrypt and human visual model. The meaningful gray watermarking image was encrypted by two-dimensional Logistic chaotic map. The block embedding intensity was calculated being combined with the human visual model so that the embedding and detection steps of encrypted binary watermark can be adaptively fulfilled in the multi-resolution wavelet coefficients of the host image. The experimental results have shown that this watermarking technique, in which the encrypting and embedding algorithms have less computational capacity, can endure regular digital image process and has good performance.

ID-based content extraction signature

2007, 27(10):  2456-2458. 

Asbtract ( )   PDF (570KB) ( )  

Related Articles | Metrics

A new ID-based Content Extraction Signatures （CES） scheme based on the batch signature was proposed. In our scheme, the signer only needs to sign the whole message. But other schemes have to sign each partition of primitive formation, so it improved the efficiency, And the dishonest Private Key Generator can not impersonate any user at any time. The proposed signature scheme can defend existential forgery on adaptively chosen message and ID attack in the random oracle model.

Signature scheme with message recovery based on discrete logarithms and factoring

Xi-Feng YUAN Yan-Rui SUN Jin-Qing SUN Ying-Hui YANG

2007, 27(10):  2459-2460. 

Asbtract ( )   PDF (589KB) ( )  

Related Articles | Metrics

Recently, there is little research about digital schemes with message recovery based on double hard problems. The computational efficiency and transmission efficiency of the existing schemes is too low. Hence, in the paper, a new digital signature scheme with message recovery was given, in which the security was based on the difficulties of computing discrete logarithms and factoring. And its security analysis and efficiency analysis were also given. The security of the scheme is consequently better than those of the signature schemes which are based on the difficulty of a single problem. And the scheme proposed has higher efficiency than the schemes which exist now.

Digital watermarking-based media automatic identification technology

2007, 27(10):  2461-2463. 

Asbtract ( )   PDF (619KB) ( )  

Related Articles | Metrics

It makes recognition simple and feasible with a computer if we embed its feature information in this entity in the form of digital watermarking, meanwhile, it will not affect its sensibility. In this article, a mode was set up to realize this idea in some particular areas, and related algorithm was also given out. And it is an algorithm with high performance, and the goal of recognition and security can be also archived effectively with it.

Secure clustering algorithm based on trust evaluation in tactical Internet homogeneous layer

2007, 27(10):  2464-2469. 

Asbtract ( )   PDF (1308KB) ( )  

Related Articles | Metrics

A kind of secure clustering algorithm Secure Clustering Algorithm Based on Trust Evaluation （SCABTE） was proposed based on trust evaluation. It was suitable in constructing the tactical Internet homogeneous layer network structure. The detailed design tactics of cluster forming and maintenance were provided. The security problem was analyzed from the point of view of invasion node. Finally, the usability of the algorithm was validated in the NS-2 simulation environment adopting Reference Point Group Mobility Model （RPGMM）.

Adaptive color image steganography algorithm based on chaos encryption

2007, 27(10):  2470-2472. 

Asbtract ( )   PDF (771KB) ( )  

Related Articles | Metrics

A new kind of design and implementation of a steganography algorithm based on Least Significant Bit （LSB） was proposed, which used chaos system as key to encrypt the steganography information, then according to the image characteristic, adaptively regulated embedding depth. The results suggest: good invisibility can be achieved, and the hiding capacity of information is large. The new steganography algorithm, based on the color image, solves the communication security problem of the information at Internet well.

Algorithm based on key to confuse watermark embedding positions

Xiang-Xiang ZHOU Zhong-Hai YIN

2007, 27(10):  2473-2474. 

Asbtract ( )   PDF (622KB) ( )  

Related Articles | Metrics

An algorithm based on the key to hash the digital watermark embedding positions was presented. The Scrambling algorithm used shifting and shuffling to hash the watermark embedding positions many times. It is very simple and convenient to carry it out, and then the positions are stochastic and dispersive. The security of embedding positions relies on a key, which to a certain extent makes the result of scrambling perfectly, and increases the security of the system.

Digital signature project based on BB84 and elliptic curve cryptography

Chun YANG Li JIAN

2007, 27(10):  2475-2477. 

Asbtract ( )   PDF (578KB) ( )  

Related Articles | Metrics

Based on the quantum key distribution protocol of BB84 and elliptic curve cryptography, a new digital signature scheme was proposed. This project used the quantum key as conversation key to make the signature process be highly effective and simple. The certifiable security of this conversation key in the key distribution process combined with the elliptic curve encryption system has provided the dual safekeeping of security to the new scheme and simultaneously achieved mutually authenticated effect.

Software process technology

Hybrid ant colony algorithm for solving TSP based on path exchanging

Dong WANG

2007, 27(10):  2478-2480. 

Asbtract ( )   PDF (554KB) ( )  

Related Articles | Metrics

It can restrain premature of ant colony algorithms and accelerate the convergence rate of the algorithms, combining ant colony algorithms with heuristic algorithms. The solution quality and efficiency of heuristic algorithms can be improved through establishing reference optimization edge set used by local search algorithms. The strategy of path exchanging can improve convergence rate and capacity of searching optimal solution. The results of experiments indicate that new hybrid ant colony algorithm can find global optimal solution of TSP whose scale is less than 2000.

Quantitative analysis method for causality knowledge based on time-event exchange diagram

2007, 27(10):  2481-2483. 

Asbtract ( )   PDF (712KB) ( )  

Related Articles | Metrics

For the void of time information and quantitative methods of the knowledge analysis for conjunction causality events, a quantitative analysis method based on time-event exchange diagram was proposed. Firstly the time-event exchange thought was adopted to establish the description mechanism that contained time information and based on event domain of causality knowledge. Then the happening and cumulate time influences of the cause event to the result event are considered. Afterwards the causality-degree, rely-degree and effect-degree including their calculating formulas are introduced which realize the quantitative description and bidirectional reasoning of causality knowledge. An example of application demonstrates that the method has high efficiency, elaborate and visual advantages.

Solving a class of combinatorial optimization problems with mixed evolutionary algorithm

Biao-xinng Pu

2007, 27(10):  2484-2486. 

Asbtract ( )   PDF (557KB) ( )  

Related Articles | Metrics

A class of combinatorial optimization problems has the characteristic that problem space can be mapped into a set of permutations. A new method to solve the combinatorial optimization problems with mixed evolutionary algorithm based on natural scale enumerative system coding was proposed in this paper. Elitism strategy, one-point crossover and Gaussian mutation were used in the algorithm. An ingenious method for operation among big parameters was designed, which can avoid overflow and simplify the operation. Simulation results and analysis show that the new method is feasible，effective and all-purpose.

Market-driven strategy for multi-agent automated negotiation

XIANG Zhao-Xia 向朝霞

2007, 27(10):  2487-2489. 

Asbtract ( )   PDF (658KB) ( )  

Related Articles | Metrics

The negotiation strategies of agents in agent-based negotiation systems are mostly static in e-commerce. To solve this problem, this article presented a market-driven strategy. Negotiation agents can make adjustable rates of concession by reacting to changing market situations, help users to make optimal trading decisions and select the appropriate strategies automatically. Experiments show that the results of using market-driven strategies are more satisfying to users than using fixed strategies.

Improved genetic algorithms based on chaotic mutation operation and its application

Xiao-Lin Gu

2007, 27(10):  2490-2492. 

Asbtract ( )   PDF (525KB) ( )  

Related Articles | Metrics

In order to change the slow convergence of simple genetic algorithms, an improved genetic algorithms based on chaotic mutation operation (named as CGA) was presented. In the evolution process, to avoid local premature convergence, CGA introduced a chaotic mutation operator in the mutation of the optimum individual, and enlarged the scope of chaotic campaign to the range of variables optimization, through the evolution from generation to generation, converged to a individual, which is the most suitable environment, to find the optimal solution; the best individual sequences is to avoid missing the best value. At last the examples were emulated, and the data make it clear that the CGA is super to SGA.

Research of model checking UML

2007, 27(10):  2493-2497. 

Asbtract ( )   PDF (970KB) ( )  

Related Articles | Metrics

Unified Modeling Language (UML) is the most commonly used method in software system design and analysis, and how to verify that the UML model satisfies some properties is a very important issue. Model Checking is an efficient automatic technique for the improvement of system's reliability, and this paper is a research into how to check UML model through the Simple PROMELA Interpreter （SPIN） model-checker. After describing the UML model using formal method, we first used hierarchical automata to express statecharts diagram. In addition, we translated the statecharts and part of the class diagram into PROMELA based on the operational semantic of hierarchical automata, and verified the model by using Simple Promela Interpreter （SPIN） to test if it satisfied the LTL properties. We also verified the consistency between sequence diagram and statecharts diagram by using LTL formula to express sequence diagram. Finally, based on the method, we developed a model checker tool called UMLChecker.

New DAO pattern with dynamic extensibility

2007, 27(10):  2498-2500. 

Asbtract ( )   PDF (477KB) ( )  

Related Articles | Metrics

Currently the existing Data Access Object （DAO） patterns have several limitations. First, the interface of the patterns and business objects is tightly-coupled, which seriously affects the dynamic extensibility of software systems. Second, the patterns have duplicated implementation codes, which adds to the difficulty of system maintenance. To solve these problems, a new DAO pattern with stronger independency and dynamic extensibility was proposed in the paper based on the concepts of data binding, meta data and meta model. Finally, an example was given to illustrate the using process of the new DAO pattern. The greatest advantages of the new DAO pattern are as follows. If any business object is needed to add to the system, we don't have to modify any codes of the class DAOFactory. All that we need to do is to modify the mapping file. Furthermore, because we have only one DAO implementation class to accomplish all the data access to business objects, if some SQL statements are needed to be modified, all we need to do is to modify the DAO implementation class but not modify any business objects.

Coloured Petri net extended with price information and its application

峰 刘

2007, 27(10):  2501-2503. 

Asbtract ( )   PDF (504KB) ( )  

Related Articles | Metrics

This paper proposed a price coloured Petri net which is a kind of coloured Petri net extended with price information to meet the demand of cost and budget analysis in the business process modeling. The transition rule of the price coloured Petri net was defined，then the analysis technique and its application were introduced in a case of business process. The conclusion is that incorporating price information in coloured Petri net is feasible and valid.

Research of Ontology-based user model

2007, 27(10):  2504-2507. 

Asbtract ( )   PDF (789KB) ( )  

Related Articles | Metrics

As a conceptualized model for the specific domain, ontology can explicitly describe concepts and their relationships, and be used for semantics expansion and inference. Currently some user models based on ontology have emerged. But there are some limitations and shortcomings. A new Ontology-based user model has been proposed, which constructs the domain ontology by text mining, and updates user model by ontology learning.

Dynamic user access control mode extending RBAC for CRM system & implementation

2007, 27(10):  2508-2511. 

Asbtract ( )   PDF (840KB) ( )  

Related Articles | Metrics

By discussing the characteristic of user management of Customer Relationship Management （CRM） System, a novel dynamic user access control model, extending the model of Role-Based Access Control （RBAC）, was brought forward in this paper. This paper gave the definition and analysis of the model in detail. Firstly, the model used User-Role-Permission mechanism to reduce the complex of authorized management. Secondly, it used the hierarchic role and separation of duty constraint to solve the overlapping problem of roles. Thirdly, it introduced the concepts of Team and Personal datasets to solve the problem of accessing data separately, and they also improved the personalized service. Finally, this paper provided a practical system instance.

Graphics and image processing

Parallel implementation of neighbor-based FCM clustering for remote sensing image

2007, 27(10):  2512-2514. 

Asbtract ( )   PDF (977KB) ( )  

Related Articles | Metrics

Considering the spatial relationship of pixels when it is used in classification for remote-sensing imagery, Neighbor-based FCM algorithm was put forward by modifying the value of fuzzy membership degree with the neighbor information during the clustering iterations. We use dominant class, if it can be determined in a fixed neighbor region, or the weighted parameters based on the distance of neighbors to perfect the membership degree of central pixel. Then parallel implement for the algorithm was also proposed by taking account of the communication complexity and the spatial relationship for image partition. In the end, the experimental data indicate the efficiency of the algorithm in decreasing the clustering iterations and increasing the classified precision, and the parallel algorithm also achieves the satisfying linear speedup.

Image denoising using nonsubsampled contourlet transform

2007, 27(10):  2515-2517. 

Asbtract ( )   PDF (701KB) ( )  

Related Articles | Metrics

To remove noise from image a method based on Contourlet Transform was proposed. The construction of Contourlet Transform is based on a nonsubsampled pyramid structure and nonsubsampled Directional Filter Banks. The threshold for coefficients shrinkage was tracked across scales and a library for scale-dependent threshold was built by performing Contourlet Transform on simulated Gaussian noise samples. Images that are corrupted with additive Gaussian noise were de-noised by coefficients shrinkage where threshold was looked up directly in library by estimating standard deviation of noise. The experimental results show that the proposed approach performs effectively in terms of both vision and the PSNR values.

Improved tMHI based real-time video moving object detecting and tracing

Qing-Shuang Zeng

2007, 27(10):  2518-2521. 

Asbtract ( )   PDF (1216KB) ( )  

Related Articles | Metrics

A new method for segmenting and tracing real-time moving object from video image was purposed, this method using the step-down grey value silhouette of timed Motion History Image (tMHI). By enveloping and marking the motion silhouette regions, it can segment the moving objects from video image real-timely. By associating the motion silhouette regions in every tMHI with the moving objects in real scene, it achieved the trace of moving objects. To get high quality segmentation, the primary tMHI was improved by removing most of the noise. The method was demonstrated with a traffic video, and some moving objects were segmented effectively. It appeared that the method has good locating robust; high detecting rate and fast processing speed that satisfied real time usage. Part conglutination was basically avoided also.

Surface reconstruction from scattered points based on Bayesian

Qi Xing Changqian Zhu Qiang Peng

2007, 27(10):  2522-2524. 

Asbtract ( )   PDF (957KB) ( )  

Related Articles | Metrics

A surface reconstruction algorithm from noisy point clouds based on Bayesian statistics was presented. The main idea is to perform a search for a maximum of posterior probability (MAP) in the space of possible constructions. First, a mathematical model of noisy measurement process and a prior over surface shapes were computed respectively. Second, an approximate MAP-reconstruction for each point was found by using a conjugated gradient optimization method. Finally, the Surface Splatting algorithm was applied to render the constructed point-based models. Experiments show that the proposed prior can smooth away noise of the scanned point clouds while enhancing visible surface features. Compared to previous work in this area, the algorithm in this paper can obtain better reconstructing results.

Shape adjust based on surface of interpolating of Doo-Sabin subdivision

ZHEN Ti-Wei

2007, 27(10):  2525-2529. 

Asbtract ( )   PDF (1084KB) ( )  

Related Articles | Metrics

The construction of an adjustable smooth surface interpolating a mesh of arbitrary topological type is an important problem in many graphics applications. The paper presented a modification of the control mesh and a subsequent Doo-Sabin subdivision. The method has the following features: 1) It can adjust extremely surface according to a scalar and can interpolate some or all vertices of a mesh with arbitrary topology, thus making the new method itself suitable for interactive free-form shape design. 2) The whole complexity is O (k), where k is the number of the vertices. In the end, we discussed shape adjust.

3D reconstruction based on fundamental matrix estimation weighted by match measure

Li-Chun LI Zhi-Qiang Qiu Kun-Pen WANG Qi-Feng YU

2007, 27(10):  2530-2533. 

Asbtract ( )   PDF (1112KB) ( )  

Related Articles | Metrics

The evaluation of correspondences was introduced to the 3D reconstruction form binoview, which was defined as match measure function and gave a guideline to use the correspondences for estimating the Fundamental Matrix (FM). The works of features matching, FM computing and 3D reconstruction became one union process. The detecting and matching of features were analyzed firstly; then two match measure functions were defined according to the correlating and feature-distance-comparing matching methods. The new linear algorithm computed FM from the correspondences weighted by its match measure which was based on the normalized 8-points algorithm, meanwhile the RANdom SAmple Consensus （RANSAC） algorithm was employed to overcome the outliers. With the internal camera parameters, the camera relative motion was computed from FM. An optimal algorithm was employed to get the accurate result At last, reconstruction was realized through triangulation. Experimental results with simulative and real images show that the algorithm works robustly with high accuracy.

Optimized recording approach of surveillance system

Wu Pei

2007, 27(10):  2534-2537. 

Asbtract ( )   PDF (1006KB) ( )  

Related Articles | Metrics

Focusing on security surveillance system, an approach about how to code different recorded object with different tactic and different precision was deduced. This paper used high precision to code the important object in the suspicious scene and, and used relatively low precision to code the other moving object. For the background scene, we only recorded one or several key frames about the background changing. Meanwhile, we deduced a fast algorithm to detect and track the suspicious motional scene which needed to be recorded and, omit the recorded frames about the quiescent scheme. Experimental results show that much storage resource could be saved and the rate assignment could be more reasonable during the coding course.

Algorithm of real-time soft shadows based on shadow maps

2007, 27(10):  2538-2540. 

Asbtract ( )   PDF (1083KB) ( )  

Related Articles | Metrics

A new algorithm of real-time soft shadows based on shadow maps was presented. Firstly, we generated two depth buffers for occluder and receiver, discretizing the occluder to micro-patches; then, we calculated the depth value of shadow for each micro-patch and estimated the visibilities of them; finally, we computed penumbra extent for each micro-patch and projected soft shadow map on the scene to obtain real-time soft shadows. Experiments demonstrate that the algorithm can provide visually pleasing effect for the complex objects and achieve satisfying frame rate at the same time. Therefore, we offered an effective way to generate real-time soft shadow for the complex scenes.

Intelligence perception

Face feature recognition based on depth information

Cheng-gan SHAN Qing-xin HU

2007, 27(10):  2541-2543. 

Asbtract ( )   PDF (825KB) ( )  

Related Articles | Metrics

A new method for positioning the face feature point by extracting depth information of partial face was proposed. Firstly, face region was determined by using clustering of the skin color in YCrCb chroma space. And then the proposed system got the relative depth of partial face through setting up a feed forward neural network to position the nose. Experimental results demonstrate that the performance of the proposed methodology is good.

Improved moving objects detection method based on Gaussian mixture model

Yi-De MA Wang-Fei ZHU Shi-Xia AN Hui-Yin QIU Shu-Sen TANG

2007, 27(10):  2544-2546. 

Asbtract ( )   PDF (965KB) ( )  

Related Articles | Metrics

This paper proposed an improved moving objects detection method based on Gaussian mixture model in the case of focusing on a video monitoring system with a static camera. First, for updating the parameters (mean and variance) of the Gaussian models, the learning rates of mean and variance were different: for mean, the learning rate was adaptive, while for variance, the learning rate was fixed; Second, The notion of Mean Of the Weight (MOW) was introduced, which had a big contribution for differentiating background points from foreground points. Third, Shadow was detected and removed with the help of background image. Experimental results show that the proposed method possesses better ability of learning and higher efficiency of detecting large and slow objects in busy environments.

Speech recognition based on two-dimensional PMCC robust feature parameter

Jin-Bao LI

2007, 27(10):  2547-2548. 

Asbtract ( )   PDF (339KB) ( )  

Related Articles | Metrics

One of the key problems in noise speech recognition is how to extract the robust feature parameters. A two-dimensional root cepstrum feature parameter was first proposed, and then this parameter was combined with robust feature based on the Minimum Variance Distortless Response (MVDR) method of spectrum estimation proposed. Finally, a novel robust feature was presented to be successfully used in continuous speech recognition under different SNRs. Experimental results indicate that the two-dimensional PMCC robust feature parameter is superior to conventional Melt Frequency Cepstral Coefficients （MFCC） and Perceptually Linear Prediction （PLP） in improving the recognition accuracy under different noise conditions and SNRs.

Improved density-based routing algorithm for wireless sensor networks

Xi-Huang ZHANG Cui-Fang GAO

2007, 27(10):  2549-2551. 

Asbtract ( )   PDF (637KB) ( )  

Related Articles | Metrics

Real energy supply is necessary for selecting routing nodes. So the energy correlation of neighbor nodes was used to reflect the local energy availability. On the basis of potential energy in the Density-Based Routing (DBR) algorithm, this paper aimed to further reduce energy consumption in the routing process. The concept of residual hop was introduced to reflect the real transmission capability of nodes. An optimized routing model was proposed with new metric. The simulated results showed that the improved routing algorithm could achieve better network performance by low-cost techniques.

Study on behavior characteristics of WSN based on minimum-hop-count routing

Ming-Cai Zheng XiaoChao Zhao

2007, 27(10):  2552-2555. 

Asbtract ( )   PDF (799KB) ( )  

Related Articles | Metrics

Repeat transmission of information and data packets is a measure to improve the reliability in Wireless Sensor Networks （WSN）, while some other performance is degraded greatly. To resolve the problem, it is necessary to make clear of the characteristics of a given wireless sensor network on running time. In this paper, with the help of theoretical analysis and simulation, the behavior characteristics of minimum-hop-count-routing-based wireless sensor networks were revealed, the factors influencing the degree of message's repeat forwarding and the relation between the factors and the packet's duplicates were given. This can provide efficient guide for ensuring and controlling the degree of the packet's duplicate transmission.

Ultra-lightweight IPv6 protocol stack of wireless sensor network

Shu-Bin SONG Neng WANG

2007, 27(10):  2556-2558. 

Asbtract ( )   PDF (876KB) ( )  

Related Articles | Metrics

According to 6LowPAN drafts, an ultra-lightweight IPv6 protocol stack was greatly tailored for wireless sensor networks by studying the tree topology of wireless sensor network, adaptation functions and IPv6 protocols. The design has dramatically reduced the codes space and the overheads of the communication among IPv6 protocols, and it meets the constraints of low energy consumption and limited processing power in wireless sensor network.

Improved image recognition and registration algorithm based on convex hull and affine invariant

Wei-Dong Zhao

2007, 27(10):  2559-2562. 

Asbtract ( )   PDF (748KB) ( )  

Related Articles | Metrics

An improved image recognition and registration algorithm based on convex hull and affine invariant was presented. The convex hulls were extracted from the feature points of the reference and test images, and then the affine invariants were calculated. These invariants were used to establish the correspondence between the convex hull vertices of the test image and the reference image in order to undo the affine transformation between them, and then a feature points matching approach for recognition followed this. This method is very well suited to handle the occasions that the object in the image has undergone an affine transformation and is partly covered.

Fingerprint singular points detection based on both weighted averaging gradient directional field and improved Poincare Index

2007, 27(10):  2563-2565. 

Asbtract ( )   PDF (542KB) ( )  

Related Articles | Metrics

Because the fingerprint is characterized by the local parallel, gradual change and high correlation of neighborhoods, a gradient based weighted averaging algorithm was proposed. The conventional poincare index algorithm for detecting the singular points of fingerprint was improved. The experimental results suggest that the improved poincare index algorithm on the weighted averaging gradient directional field can accurately detect the singular points of fingerprint.

Typical applications

Applied research of clustering algorithm in telecom consumer segments

2007, 27(10):  2566-2569. 

Asbtract ( )   PDF (867KB) ( )  

Related Articles | Metrics

With the analysis of clustering algorithms, an applied model for telecom customer segments was provided, which achieved good applications in actual customer market segments, and provided reference for designing telecom service products. With introduction to definition of index difference, a new method was presented to evaluate the application effect of clustering algorithms. The method shows the best performance of K-means for actual cases in telecommunication customer segments with multiple clustering algorithms such as K-means, SOM, and BIRCH.

ISOMAP based on minimal connected neighborhood graph

Chao Shao

2007, 27(10):  2570-2574. 

Asbtract ( )   PDF (958KB) ( )  

Related Articles | Metrics

It is well known that ISOMAP is poorly robust and topologically unstable, mainly because “shortcut” edges may emerge in the neighborhood graph due to the noise or the unsuitable neighborhood size. The emergence of “shortcut” edges can make the corresponding neighborhood graph represent the neighborhood structure of the data falsely, and thus ISOMAP cannot be applied successfully. Therefore, this paper presented a more robust and more topologically stable ISOMAP algorithm, i.e. MCNG-ISOMAP (Minimal Connected Neighborhood Graph-based ISOMAP), which can prune effectively "shortcut" edges, existed possibly in the neighborhood graph, based on that the minimal connected neighborhood graph can avoid "shortcut" edges effectively. MCNG-ISOMAP is much less sensitive to the neighborhood size and thus can be applied to data visualization more easily than ISOMAP. In addition, MCNG-ISOMAP can also find the true intrinsic dimensionality of the data independent of the neighborhood size unlike ISOMAP. Finally, the feasibility of MCNG-ISOMAP is verified by experimental results very well.

Study on the errors in short time spectrum analysis of speech signal and the all phase DFT spectrum

2007, 27(10):  2575-2577. 

Asbtract ( )   PDF (484KB) ( )  

Related Articles | Metrics

This paper investigated the window effects and the spectrum amplitude errors in hamming windowing spectrum analysis in detail and gave its compensation coefficients. Simulation results show that these coefficients are correct. Furthermore, this paper introduced the all phase DFT spectrum analysis and tested its good performance in improving the resolution of spectrum in speech signal analysis.

Grid-based LiDAR data processing platform architecture

2007, 27(10):  2578-2580. 

Asbtract ( )   PDF (839KB) ( )  

Related Articles | Metrics

The software architecture in Light Detection And Ranging （LiDAR） data processing was described. It partitioned LiDAR data and processed task into distributed Grid nodes. Processing speed was improved by parallel computing of every Grid node. The Open Grid Service Architecture （OGSA） also provided flexible share style for LiDAR data and processing service.

Optimization research of spatial index structure of R*-tree

min chen

2007, 27(10):  2581-2583. 

Asbtract ( )   PDF (580KB) ( )  

Related Articles | Metrics

Aiming at the need of very large spatial database applications and taking notice of the disadvantagesf existing spatial indexes, the thesis introduces an improved R*-tree spatial indexing structure after addressing the related concepts, data structures, algorithms, and performannalyses of R-tree and R*-tree. This paper gave a more efficient algorithm which analyses the shortcoming of R*-tree spatial index algorithm. The experiment results show that improved R*-tree is better than the original R*-tree in index efficiency.

Using Gini-Index for feature selection in text categorization

Yong-Min LIN Wei-Dong ZHU

2007, 27(10):  2584-2586. 

Asbtract ( )   PDF (720KB) ( )  

Related Articles | Metrics

This paper used improved Gini-index for text feature selection, and constructed the measure function based on Gini-Index， then compared it to other four feature selection measures using two kinds of classifiers on two different document corpora. The result of experiments shows that its performance is comparable with other text feature selection approaches. However, it is perfect in the time complexity of algorithm.

Data preparation model based on hierarchical road network and its implementation

2007, 27(10):  2587-2590. 

Asbtract ( )   PDF (747KB) ( )  

Related Articles | Metrics

To reduce the communication time, we proposed a data preparation model named transportation meshes set circum-rectangle（TMSCR）. This modal stated that traffic center should send a special data set to vehicles, which consisted of the low level road segments in O-grid and D-grid as well as high level road segments in the circum-rectangle of meshes set which contained O-node and D-node. Experiments indicate that the implementation process is much more efficient in a personal computer. The model will greatly contribute to time savings in the communication between traffic center and vehicles so that provide better service level to vehicles compared with traditional method.

Real time detection and classification algorithm for abandoned and stolen objects

WANG Wei-Jia 王伟嘉 Hui LIU Li SHA LIU Xin JIANG Hua

2007, 27(10):  2591-2594. 

Asbtract ( )   PDF (1124KB) ( )  

Related Articles | Metrics

Algorithm for real time abandoned/stolen objects detection by a single and static camera was studied. It intended to solve the problem that the detection validity of gradient-based method will decrease under the condition that the gradient of environment is complex. Besides the edge points' position analysis of contour, the edge points' connectivity analysis of contour was introduced. Only the object of which the contour reaches the similarity of both the edge points' position and connectivity can be recognized as the abandoned object. In addition, Histogram-based method which calculates and compares the Bhattacharya distance between histograms to determine if the object is abandoned or stolen was studied, and was compared with the first method. Experimental results show that the improved gradient-based method is more adaptable and effective.

Research and design of location service architecture in migrating workflow system

Yu-Feng QIN Guang-Zhou ZENG

2007, 27(10):  2595-2597. 

Asbtract ( )   PDF (801KB) ( )  

Related Articles | Metrics

The services to Migrating Instance （MI） that a location can be offered were classified into MI services and workflow services at first, and then a service architecture based on multi-thread to support these two kinds of services was proposed in this paper. It not only helps to increase the services efficiency of the location server provided, but also makes it more extensible.

Algorithm of two-phase approximate KNN outliers mining and its application

2007, 27(10):  2598-2601. 

Asbtract ( )   PDF (839KB) ( )  

Related Articles | Metrics

In this paper, we targeted at high-dimensional datasets and presented Two-Phase Approximate KNN Outliers Mining （TPOM）, an algorithm of two-phase based approximate KNN outliers mining. It processed clustering on the datasets firstly, and then improved the efficiency of nested loop KNN outliers detecting by accelerating the nearest neighbors search and improving the cut-off efficiency. Application and the result show that TPOM is well suit and extendible to real data, it scales log-linearly as a function of the number of data points and linearly as a function of the number of dimensions.

New score normalization method in speaker verification

Xin-Jian Gao Dan Qu

2007, 27(10):  2602-2604. 

Asbtract ( )   PDF (529KB) ( )  

Related Articles | Metrics

In the speaker verification field, the distributions of the target and imposter scores are bimodal, and the output score distributions of different target speaker models are different too, so it is difficult to estimate a common threshold. The most frequently used score normalization techniques: T-Norm and Z-Norm were analyzed. A new technique D-Norm was introduced based on the KL distance. Then a new method ZD-Norm was proposed based on Z-Norm and D-Norm. The experiments show that the ZD-Norm method can improve the performance efficiently and make the estimation of threshold easier.

Signal de-noising in wavelet based on new threshold function

ZHU Hua 朱华

2007, 27(10):  2605-2606. 

Asbtract ( )   PDF (510KB) ( )  

Related Articles | Metrics

A new threshold function based on 3σ-rule and SemiHard Thresholding （SHT）,SemiHard thresholding was presented. It overcame the shortcoming of the hard threshold with discontinuous function and solved the problem of the permanent bias in soft threshold function. The results show the de-noising effects are improved.

Personalized text content filtering based on typical feedbacks

Juan-Li HU

2007, 27(10):  2607-2609. 

Asbtract ( )   PDF (574KB) ( )  

Related Articles | Metrics

This paper presented an approach to perform text content filtering based on content-based and collaborative filtering, using the Probability Model. Introducing the idea of Stereotypic inference, it classified the users into different types and built the model for each type. Moreover, the refreshing of the profiles was based on the feedbacks of the model, and the building of the new profiles was based on the typical profiles. In this way, the precision and the recall were improved significantly.

Pedestrian abnormal behavior detection based on region feature

2007, 27(10):  2610-2611. 

Asbtract ( )   PDF (570KB) ( )  

Related Articles | Metrics

Abnormality of pedestrian was detected based on region feature in this paper. The principle of the method is simple but effective. Firstly, moving object was detected with mixture of Gaussians background model and the minimal rectangle of the object whose ratio of height to width was taken as the feature to measure abnormality of the pedestrian. Experiments demonstrate the approach proposed is fast and effective.

Implementation of embedded real-time driving fatigue monitoring system based on DM642

2007, 27(10):  2612-2614. 

Asbtract ( )   PDF (710KB) ( )  

Related Articles | Metrics

The major frame of the real-time embedded driving fatigue monitoring system based on DM642 was discussed. A method of eye detection was presented by using controllable LEDs and differences of odd-even frames. In this method, the odd and even frames were captured respectively through synchronizing the inner and outer ring of LEDs. Because of the red-eyed effect generated by adaxial LEDs, one frame had bright pupils and the other had dark pupils. Therefore, the pupils were easy to be detected and tracked from the difference of the two frames. Then the characteristic parameters of eyes could be analyzed. After that, the closing time of eyes within regular successive time was counted. Finally the percentage of eyelid closure over the pupil over time （PERCLOS） value was calculated, and the tiring degree of driver could be judged and alarmed when necessary. Experimental results show that the fatigue detection rate in this system is high, and it can be used to monitor the driver's fatigue state.