Abstract:CC4 (the 4th version of corner classification) neural network is a new type of corner classification training algorithm for three-layered feedforward neural networks. It has been provided as a document classification approach for metasearch engine Anvish. On the condition that documents are almost of the same size, CC4 neural network is an effective document classification algorithm. However, when there is great difference in document sizes, CC4 neural network does not perform well. This paper aims to extend the original CC4 neural network for effectively classifying documents having much difference in sizes. To achieve this goal, the authors propose a MDS-NN based data indexing method thus making all documents be mapped to k-dimensional points while their distance information is kept well. The authors also extend CC4 neural network so that it can accept k-dimensional indexes of documents as its input, then transform these indexes to binary sequences required by CC4 neural network. The experimental results show that the performance of ExtendedCC4 is much better than that of InitialCC4 when there is a great difference in document sizes. At the same time, the high classification precision of ExtendedCC4 has much relationship with the effectiveness of indexing methods.

Abstract:Palmprint recognition aims at finding out the palmprint template from the database, which is from the same palm as a given palmprint input. Feature extraction plays an important role in the recognition process. In this paper, we propose a new feature extraction method by converting a spatial domain palmprint image into a frequency domain image using Fourier Transform (FT) and representing palmprint features in the frequency domain. The extracted features are used as indexes to the palmprint templates in the database and the searching for the best match is leaded by these features in a layered fashion. The experimental result shows that the proposed FT based feature extraction method is effective in terms of accuracy and efficiency on our palmprint database.

Abstract:Mobile agents refer to self-contained and identifiable computer programs that can migrate along the network and can act on behalf of the user or other agents. Mobile agents often work on heterogeneous network and operating system environment. Therefore, an integrated logical interface to access physical structure via mobile agent application has become more and more urgent. In this paper, we proposed a tracking and persistent agent-based mobility management system in case of distance learning. The main purpose of our system is addressed to achieve the universal access objective. In order to let the whole mobility management system full play, firstly, we proposed a mobile agent communication network model and the mobile agent evolution states. The mobile agent communication network model facilitates the agent computation, and the mobile agent evolution states assist agent supervisor to monitor client agents. Secondly, we encapsulate the utility tools to be a role-setting object. The role-setting object is an application-driven component, which can provide customization benefits for user to match user's demands. These integrated technologies are sufficient in distance learning (virtual university) environment.

Abstract:Asynchronous Transfer Mode (ATM) is an emerging standard for broadband networks, packetized video is likely to be one of the most significant high-bandwidth users of such networks. The transmission of variable bit-rate (VBR) video offers the potential promise of constant video quality but is generally accompanied by packet loss, delay and delay jitter which significantly diminishes this potential, so it is necessary to provide requested QoS. In this paper, two new dynamic bandwidth allocation algorithms for MPEG videos over ATM are proposed, one is dynamic measure algorithm, another is dynamic virtual buffer algorithm, in which the authors dynamic adjust the bandwidth of MPEG video sources with a specified allowed cell loss ratios to provide QoS guarantees in an ATM network.

Abstract:A novel method to segment image is proposed considering the intensity roughness and the color purity. HIS color space is used to express pixel color information. Not only the pixel properties but also the pixel group properties are considered. An image is divided into blocks. For each block, the mean and variance values can be seen as the pixel group properties. Because the hue value will be very sensitive when the intensity and saturation values are small, the color purity is used as the weight on the hue value. The image segmentation is done with the approach of the combination of region growing based on block and vector quantization. The experiments confirm this method is suitable for many kinds of images.

Abstract:To specify and verify fault-tolerant systems in XYZ/E is discussed in this paper. Based on the corresponding state transition system of an XYZ/E executable program P, how to model its fault environment and obtain its fault affected program P_F by fault transformation is illustrated. With P, F, P_F and a recovery algorithm R, how to obtain the fault-tolerant program P_F-R by fault tolerant transformation is also illustrated. Furthermore, two kinds of refinement relationships between programs P and Q: fault-tolerant refinement and backward-recovery refinement are defined.Based on these two refinement realtionships,some properties satisfied by program Q can be directly deduced from the specification of programP.

Abstract:With the distributed object technology becomes a hot topic, more and more developers begin to use ORBs as the applications?basis of developing and executing. ORB can simplify the development of large application by facilitating the lower-level communication between objects. But from the view of developers, this kind of support is not enough yet, such as the introduction of too many new concepts (especially those related with ORB), and the incompatibility between the ORBs from different vendors. These proble bring difficulties to the application developers for their mastering the ORB and make the applications portable between the different ORB p;atforms.After analyzing the issues in the ORB-based application development,an automated approach is presented to improvr and facilitatr the development process.The prototype of the supporting tool shows that by generating skeleton codes and separating ORB product related codes,the proposed method can efficiently reduce the coding work and the number of errors.

Abstract:For realizing quality of service guarantee, all multi-user networks adopt the approach of reserving resource per connection requirement, but this will induce many resource pieces in some parts of network, and traditional single routing approach cannot utilize these resources for new connection. In order to improve resource utilization ratio, a multi-path routing algorithm is proposed, which probes and reserves resources in one of several parallel paths according to the status of network and the connection Qos requirements.For verifying the validity.the proposed algorithm is implemented in the simulator of ns3 and its performances is compares with some algorithms.Under the circumstance of the network overload,can combine with several paths to admit new connection,so it improves the network resource utilization ratio and throughput.In conclusion,applying the proposed algorithm to connection-oriented mulrimedia data network will improvethe capability of transmitting multimedia information.

Abstract:Data synchronization is one of the most important methods to conciliate conflicts and maintain the data consistency between mobile clients and servers in mobile database systems. The novel concept of data synchronization object is used in Lite/2, a prototype of the mobile database system, to define a transactional synchronization method and is able to maintain the semantic integrity. Synchronization recovery is an important technique combined with the queue mechanism in Lite/2 to optimize the communication costs.The experimental results show that theLite/2 has good performance in bothsynchronization andcommunication.

Abstract:This paper focuses on two key problems in automated fingerprint identification system (AFIS): fingerprint image enhancement and minutia matching. A fingerprint image enhancement algorithm is proposed based on orientation field. In addition, a minutia matching algorithm is proposed which modified the algorithm presented by Jain et al. About this algorithm, a simpler alignment method is used. Ridge information is introduced into the minutia matching process in a simple but effective way. One of the advantages of doing so is to solve the problem of refence point pair selection with low computationl cost.Iakkition,the changeable sized boundary box is used to make this algorithm mote robust ti nonlinear deformation between fingerprint images.Experiments are done on the FVC2000 databases with the FVC2000 performance evaluation method and the results show that this improved algorithm is better than the originalone.

Abstract:A new easy technique for calibrating a camera based on circular points is proposed. The proposed technique only requires the camera to observe a newly designed planar calibration pattern (referred to as the model plane hereinafter) which includes a circle and a pencil of lines passing through the circle抯 center, at a few (at least three) different unknown orientations, then all the five intrinsic parameters can be determined linearly. The main point of the proposed technique is that it does not need know metric measurement on the model plane and the corres pondences between points on the model plane and image one,hence it can be done fully automatically.The proposed technique is particularly use for those people who are not familiar with computer vision.Experiments with simulated datawell as well as with real images shoe that the new technique is robust and accurate.

Abstract:BDI (belief, desire, intention) is a theoretical model based on agent computing, while its semantic is the key to the development of agent computing. The new BDI semantic is presented to have an evolution character by dividing B into subjective belief and objective belief, and by regarding possible world as a temporal stage of recognition procedure, which can describe agent and avoid the logical omniscience problem.

Abstract:Constraint networks provide a useful framework for the formulation of many problems in computer science. In general, constraint satisfaction problems are NP-Complete. Many problems specially restrict the structure of constraints or the form of the constraints, then allow them to be solved efficiently. For the identification of such tractable constraints, much work has been devoted to the study of special classes of constraints or constraint networks. As pointed out by Deville et al., a class of connected row-convex(CRC)constraints is shown to be tractable.This paper intrnds to finds to fint an efficient recognition algorithm for the class of constraint networks.In this paper,a standardized form for the CRC constraint matrix is proposed based on the related finding on CRC constranint network.The basic characteristics of the standardized form are analyzed,and a fast algorithm is provided for the recognition of CRC constraint networks based on a recognition algorithm of the RC constraint network,properties of PQ tree(a tree composed by P nodes and Qnodes)and an indexed matrix representation of constraints.The time complexity of the algorithm is O(n³d²),where n is the number of variables in aconstraint network and d is the maximum size of the domain for wach variable.It reaches the optimum time complexity for a CRC constraint network recognition algorithm,and hence provides afeasible solution to practical CRC constraint satisfaction problems.

Abstract:Though the loss factors in FALVQ algorithms are defined to fuzzy membership functions, the performance of the algorithms is not stable due to their scaling functions not being fuzzy membership functions. In this paper, a new family of fuzzy algorithms for the generalized LVQ network, called soft competitive algorithms for LVQ (SCALVQ), is derived from extending the definition of the loss factor corresponding to the winning prototype in FALVQ. Meanwhile, three concrete types of SCALVQ are given. In SCALVQ, the loss factors and the corresponding scaling function are both fuzzy membership functions,but an identical fuzzy membership function.Therefore,they absorb advantages of FALVQ and the soft-competition scheme,and overcome the disadvantages of the FALVQ.

Abstract:Monocular hand gesture recognition systems usually model a human hand as a pixel or a blob by which the motion of the whole hand are analyzed and the appearance features are extracted. But the system presented in this paper begins with the motion of a hand's local parts. Firstly by fusing on multiple information including motion, color and edge, the characteristic curves that can represent the structure of a hand are obtained. The characteristic curves are cut into short segments with the equal length, which,which are easily analyzed and and tracked.Then the planar model is adopted to model the appearance change between the consecutive images and calculated by motion of the short segments.At last,the deviation caused bu the coordinate system is also analyzed and a moving coordinate system is set up,from which translation-independent parameters of the planar model are exrtactrd for hand gesture recognition.

Abstract:On the basis of requirements description model HRFSM (hierarchical finite state machines based on rules), a DEM(dynamic execution model) of embedded real-time systems software and a checking method based on DEM are presented in this paper. Because DEM can integrate with control flow, data flow and time, the checking method can check consistency and completeness of software requirements for embedded real-time systems software. The checking method consists of three forms that checking purposes are different and can provide some important checking information in the checking procking procedure.Analysts can check check software requirements of embedded real-time systems soffectively by using checking tools based on the proposed method so that efficiency of analyzing and checking software requirements can be improved.

Abstract:Based on Harn抯 digital signature scheme and zero-knowledge proof, an authentication access control scheme for information protection system is presented in this paper. The scheme is safer than previously proposed one. In the scheme, two-way authentication may be done between a user and the system without exposing their secret information, and their sharing secret is used for authenticating the requesting user not to illegitimatimately access the protected file. The scheme can perform the access operation in dynamic environments,such as change access privileges and insert/delete users or files without implicating any user's secretkey.

Abstract:A dynamic (t,n)-threshold secret sharing scheme based on one-way function is proposed in this paper. It has the following properties: (1) The dealer can renew system secrets without renewing the shadows of the participants; (2) When some participants's shadows are revealed, they can be renewed without any effect on the others; (3) A new shadow can be generated for a new participant without any effect on the others; (4) The shadows can be reused for many times; (5) Only n+1 parameters should be public (When a cheater could be checked out,it should be opened 2n+1 parameters.);(6)The system secret can be recovered with a parallel process.

Abstract:In this paper, the concept of object assembling and automatic maintenance of its association semantics are discussed. After presenting the model of object assembling, the mechanism of its language description and implementation are described, which is based on object hook interfaces, rather than the ordinary object interfaces, and through which the domain-oriented dynamic manipulation of object association and automatic maintenance of association semantics are achieved. This has greatly enhanced the description ability and the dlexibility,and decreased the complexity of object association.As a mechanism of objects and object relation description and manipulation,it gives a new idea to realize software IC,and its usage,formalization and standardization will provide a new technique foe industrialized production of the software.

Abstract:In order to control the grave interference in mobile ambient (MA), Levi et al. proposed mobile safe ambients (SA). However, the coactions introduced in SA brought new security breaches. In this paper, robust ambients (ROAM) is proposed to eliminate those security breaches. In ROAM, coactions are still utilized to control the grave interference. In addition, the parameter of every coaction is explicitly specified to name the consumer of that coaction. This mechanism effectively eliminates the security breaches in SA.THe firewall crossing example and the encoding of polyadic asynchonous π-calculus in ROAM show that ROAM still keeps the strong expressiveness of its ancestors.A fundamental tupe system for ROAM with both thread count and mobility attibutes is also proposed and proved.The result in this paper shows that ROAM is a good candidate in the formalization of mobile computation.

Abstract:The support vector machine is a classification technique based on the structural risk minimization principle, and it is also a class of regression method with good generalization ability. In this paper, a new idea that each regression problem can be changed into a classification problem is presented. The proposed method has some theoretical foundations. Compared with SVM regression method, the geometric meaning of optimization problem in this paper is very clear and obvious.