15 January 2013, Volume 39 Issue 1
-
Select all|
-
Computer Engineering. 2013, 39(1): 1-6. https://doi.org/10.3969/j.issn.1000-3428.2013.01.001
To improve scientific and accuracy of disaster recovery capability evaluation, combined with disaster recovery GB, this paper proposes the resource-based disaster recovery capability index system. At the same time, for the lack of systematic research on disaster recovery capability index measurement, according to the index properties, it puts forward dynamic index measurements of threshold logic tree impact analysis based, network performance impact analysis based, and Weibull analysis based. It analyzes the application of the proposed index system and its measurement in the evaluation. The results show that the index system and its measurement specify GB layer by layer and introduce four types of dynamic index to monitor the changes in the influencing factors, therefore, to improve operability of GB and accuracy of evaluation effectively.
-
Computer Engineering. 2013, 39(1): 7-11. https://doi.org/10.3969/j.issn.1000-3428.2013.01.002
This paper proposes an Automatic Color Equalization(ACE) algorithm based on human eye lateral inhibition to solve the problems emerge after using traditional ACE, such as, poor local result, low speed. The algorithm simulates the human eye lateral inhibition mechanism, and selects the two-dimensional circular region around the target point that satisfies accords with the effective scope of the lateral inhibition. And then it selects the sampling point that obeys the Gaussian distribution. At last, it makes a differential comparison between the target point and sampling point to enhance contrast. The proposed algorithm reduces complexity as the operating is centralized among the pixels with high correlation. Experimental results show the proposed algorithm has a good details enhancement and faster processing speed.
-
Computer Engineering. 2013, 39(1): 12-17. https://doi.org/10.3969/j.issn.1000-3428.2013.01.003
Current cache Domain Name System(DNS) servers can not resist continuing Kaminsky DNS cache poisoning, so this paper proposes a defense strategy based on response packets checking. Probability theory is used to analyze the internal relation between success probability and continuing time of poisoning, which attests the harmfulness of continuing Kaminsky poisoning. Packet checking suppresses success probability’s accumulative effect on time on the existing basis so that it can be used to defense continuing Kaminsky poisoning. Simulation experiment is conducted based on probabilistic model checking tool PRISM, whose results prove that the strategy can make poison attack more difficult by over 3 600 times than it is now.
-
Computer Engineering. 2013, 39(1): 18-22. https://doi.org/10.3969/j.issn.1000-3428.2013.01.004
The problem of Narrow Band Interference(NBI) and both Multiple Access Interference(MAI) and Intersymbol Interference(ISI) in dispersive channel of the cellular mobile communication under urban architecture’s influence is proposed, a technique for the simultaneous suppression of NBI, i.e., multitone interferers, autoregressive interferers and digital interferers, and both MAI and ISI, and environmental noise is considered. This technique is based on the suboptimal subspace blind Minimum Mean Square Error(MMSE) detector of multiuser detection with the auxiliary vector to make it low-complexity. It is seen that when SNR is 10 dB, BER of blind MMSE detector is lower than about 3 dB of the conventional ways’.
-
Computer Engineering. 2013, 39(1): 23-28. https://doi.org/10.3969/j.issn.1000-3428.2013.01.005
To evaluate the invulnerability of directed and weighted complex network effectively, considering the impact of edge’s directionality and weighted to the topology of complex network invulnerability, the algorithm which is used to measure the invulnerability of directed and weighted complex network is proposed——Invulnerability Measure Algorithm of Directed and Weighted Complex Network(IMADW). This algorithm uses the shortest harmonic distance of two nodes and the whole network node to measure the nodes tightness, uses the node loop coefficients to reflect the optional paths, and gets gobal invulnerability measure value of network topology layer. The proposed algorithm is proved accurately through the algorithm analysis and experiment simulation.
-
Computer Engineering. 2013, 39(1): 29-34. https://doi.org/10.3969/j.issn.1000-3428.2013.01.006
Continuously verifying and reasoning on software’s execution property are notoriously hard to solve by general software test and static program verification. Aiming at the problems, this paper proposes a runtime program verification framework named RPA based on program instrumentation and Boolean logic. RPA defines a dynamic logic language which describes and asserts runtime properties of target program, proposes an automatic instrumentation approach to collect state information in runtime program, and designs a scheduling algorithm to support fast verification and reasoning. Experimental results show that combined with appropriate predicate extensions, RPA can effectively verify and analyze the software logic, to identify potential software errors.
-
Computer Engineering. 2013, 39(1): 35-40. https://doi.org/10.3969/j.issn.1000-3428.2013.01.007
This paper proposes a multiple satellites imaging scheduling strategy based on Integrated Index Petri Net(IIPN) and hybrid ant colony algorithm. Index information dealing with concurrent observation of multiple satellites and conflict of using satellite resource are introduced into transition, which can also reflect energy and memory constraints, making the problem description more visual and complete. An ant colony algorithm combining with local search is designed to resolve the problem. Index information is integrated in heuristic information for guiding the ants for global search. Local search technique is to accelerate convergence. Experimental results show that the method of can solve the multiple imaging satellites scheduling problem effectively by reaching balance between global search technique and local search technique which is to accelerate convergence.
-
Computer Engineering. 2013, 39(1): 41-44. https://doi.org/10.3969/j.issn.1000-3428.2013.01.008The polygon data interior point automatic generation algorithm based on Minimum Bounding Rectangle(MBR) easily leads to failure in the singular case. In order to solve the problem, this paper proposes an improved polygon data interior point automatic generation algorithm which introduces the vector data uncertainty interval concept. The improved algorithm uses the approach of uncertainty interval and intersection interval to fix the singular case, which avoids too much exception handling and branch structure of the MBR algorithm when the cutting line intersects with the node. Comparative experimental results prove the robustness and efficiency of the improved algorithm.
-
Computer Engineering. 2013, 39(1): 45-48. https://doi.org/10.3969/j.issn.1000-3428.2013.01.009It is required to translate conjunction normal form into disjunction normal form when using discernibility matrix to compute all reducts of an information system. But there exists the combination explosion problem in this transformation process. Following the properties of conjunction normal form, And operator and OR operator, this paper constructs a parallel model for the transformation process, and then proposes a multi-thread technology-based algorithm for generating disjunction normal forms. The proposed algorithm improves the efficiency of transformation by using the functions of automatic thread scheduling of Windows. Experimental results show that the algorithm seems more efficient in the cases where conjunction normal forms have relatively large length, making good use of multi-core CPU resources of computers.
-
Computer Engineering. 2013, 39(1): 49-53. https://doi.org/10.3969/j.issn.1000-3428.2013.01.010In order to improve the test efficiency, this paper proposes a unit test method based on annealing Genetic Algorithm(GA). It makes the genetic operators associate to the evolution generation, thus it can lead to faster convergence to the optimal solution. When constructing the fitness function, it uses the method based on the branch distance. It makes the decision node into parallel structure by using the code conversion function. At last, every branch distance can be quantized to the number between zero and one. Experimental result proves that the algorithm can use shorter time to achieve higher code coverage.
-
Computer Engineering. 2013, 39(1): 54-57. https://doi.org/10.3969/j.issn.1000-3428.2013.01.011According to the characteristics of Petri net can dynamically describe and analyze a system, a security required model of the railway interlocking software based on Petri net is built, then a generation algorithm of the Minimal Cut Sets(MCS) based on Petri net failure tree and dynamic generation of safety test case based on formal failure tree minimal cut sets are presented. The virtual station test results show that this method can effectively reduce the resource consumption, improve the efficiency of software testing and finally provide test basis for the interlocking software design and improvement.
-
Computer Engineering. 2013, 39(1): 58-62. https://doi.org/10.3969/j.issn.1000-3428.2013.01.012To solve recommendation problem for cold-start users, this paper proposes a collaborative filtering algorithm CoTrustWalker that combines item-based collaborative filtering and trust-based recommendation, which improves TrustWalker algorithm in similarity calculation, the possibility of item selection and prediction score. CoTrustWalker algorithm adopts the cloud model similarity method to compute the item similarities, so as to enhance the stability of recommendation results by selecting the most similar polymerization results of some items. Experimental results show that CoTrustWalker algorithm improves the quality of recommendation and the speed compared with TrustWalker algorithm.
-
Computer Engineering. 2013, 39(1): 67-70. https://doi.org/10.3969/j.issn.1000-3428.2013.01.014This paper presents an architecture-neutral locality quantization methodology, and employs it to analyze the locality of SPEC2000 benchmarks and their data, text, heap and stack. A 2-D locality distribution along axes of spatial and temporal locality intuitively exhibits benchmarks’ locality. Experimental result shows that heap section has the worst locality, which mainly constitutes the data locality of the program, while stack section has the best.
-
Computer Engineering. 2013, 39(1): 71-75. https://doi.org/wangdong@fudan.edu.cnThis paper proposes an entity search method based on multi-perspective association model for the problem of searching particular type of entities in information retrieval field. The method employs Named Entity Recognition(NER), text vector, association rules, etc, and traditional search engines as well as Wikipedia, Stanford NER etc. Experimental result on the large Web data collection provided show that, compared with BM25 and traditional Bayesian model, this method increases nDCG@R by 11.49% and 18.09% separately.
-
Computer Engineering. 2013, 39(1): 76-79. https://doi.org/10.3969/j.issn.1000-3428.2013.01.016As some anchor texts of webpage can not provide sufficient information, the performance of filling the structured database with semi-structured information of Web pages is not good in the experiments of W2DR method. This paper proposes a method based on the link path bag to integrate URL attributes. Through integrating anchor texts with Web titles, the URL attribute values are obtained and filled into the destination database from the searched Web page datasets according to the best matching strategy. Experimental results show that the F value of this method increases by 13.91% and 3.54% in two different datasets respectively.
-
Computer Engineering. 2013, 39(1): 80-84. https://doi.org/10.3969/j.issn.1000-3428.2013.01.017Aiming at the problem of API feature obfuscation caused by API noise and API rearrangement which traditional System Call Dependence Graph(SCDG) can not eliminate, this paper presents a malicious code similarity analysis method based on API dependence. The method uses a program behavior description based on SCDG composed of control dependence and four types of data dependence between APIs. API noise and API rearrangement are eliminated through data dependence analysis and control dependence normalization. Experimental results show that, compared with API sequence similarity analysis method, this method significantly improves the accuracy of the similarity analysis of malicious code.
-
Computer Engineering. 2013, 39(1): 85-88. https://doi.org/10.3969/j.issn.1000-3428.2013.01.018Aiming at the energy consumption evaluation for embedded software while running, this paper discusses an energy consumption measurement method which is based on LabVIEW and NI multifunction data acquisition card. The method is directly to acquire the data of voltage and current channel and a digital channel that is used to identify the state. The measured program is encapsulated as unit that sends state to digital channel by GPIO ports. That state determinates which program snippet a sample belongs to. An approximate calculation method is presented to evaluate the energy consumption. Energy consumption measurement result through real platform shows that the max difference is about 0.2 mJ which indicates the accuracy and effectiveness.
-
Computer Engineering. 2013, 39(1): 89-92. https://doi.org/10.3969/j.issn.1000-3428.2013.01.019This paper proposes the concepts of the MDR codes and near MDR codes over the ring F2+vF2 based on the MDR codes and the near MDS codes over field Fq, including almost MDR codes. Using a homomorphism q of an additive group from F2+vF2 to F2 and the generator matrices of the corresponding linear codes, it proves that a linear code C over F2+vF2 is a MDR or near MDS if and only if q(C) over F2 is a MDR or near MDS. Based on above results, the construction method and the examples of the MDR or near MDR over F2+vF2 are given.
-
Computer Engineering. 2013, 39(1): 93-96. https://doi.org/10.3969/j.issn.1000-3428.2013.01.020A joint preprocessing method in distributed Coordinated Multi-point(CoMP) systems is proposed so as to further improve spectrum utilization and edge user’s receptivity, and it is adopted the style of joint-coding in which pre-coding and space time block code is combined to use. According to channel state information, some relevant Remote Antenna Units(RAUs) are chosen in Distributed Antenna System(DAS) to form a cooperating set for transmission. Precoding rules are adopted correspondingly by each RAU in cooperating set, and the processed data is transmitted by using Space Time Block Code(STBC). A Maxima-ratio Combining(MRC) approach is adopted to calculate the received signals which are transmitted jointly by each coordinated RAU in this system. Simulation experimental results show that the proposed method can effectively improve not only the Bit Error Rate(BER), but also the transmission reliability.
-
Computer Engineering. 2013, 39(1): 97-101. https://doi.org/10.3969/j.issn.1000-3428.2013.01.021This paper builds an invulnerable dynamic evolution model of the network, it uses Highly Optimized Tolerance(HOT) theory and the nodes preference attachment mechanism, makes the invulnerability and self-recovery capabilities etc. of the system evolve toward the optimal direction. The node can adjust its properties to generate the invulnerable dynamic evolution model with power law distribution of node degree. Simulation analysis shows that the HOT model has a better invulnerability than that of BA model.
-
Computer Engineering. 2013, 39(1): 102-105. https://doi.org/10.3969/j.issn.1000-3428.2013.01.022In order to satisfy the requirement of energy constrained in Wireless Sensor Network(WSN), an energy-efficient multihop broadcast synchronization algorithm based on double broadcast domain is proposed. The algorithm is based on Receiver-receiver Synchronization(RRS). Node synchronization range is expanded to two broadcast domains by peer-to-peer broadcasting message in a pair of synchronized nodes which need to broadcast message one more time. A hierarchical structure with two root nodes is built, and the algorithm is extended to the network-wide. OMNeT++ simulation results show that the proposed algorithm requires much less timing messages for network-wide synchronization than MBS algorithm.
-
Computer Engineering. 2013, 39(1): 106-110. https://doi.org/10.3969/j.issn.1000-3428.2013.01.023To address the control channel saturation problem of Cognitive Radio Network Medium Access Control(CRN-MAC) protocol, a modeling scheme for control channel is proposed. The concept of Control Channel Threshold(CCT) is introduced, and according to the time constraints of superframe, control subframe and data subframe, a CCT model with computational equation is presented. The theoretical throughput of Control-Channel-based MAC(CCMAC) protocol is evaluated on the basis of the proposed model. Simulation results demonstrate the effectiveness of the model, and the network throughput is analyzed while changing parameters of the CCMAC protocol.
-
Computer Engineering. 2013, 39(1): 111-115. https://doi.org/10.3969/j.issn.1000-3428.2013.01.024In order to reduce the algorithm complexity of the target coverage problem in Wireless Sensor Network(WSN), three approaches are designed to decrease the network scale. Of the three approaches, one is deleting the redundant sensor nodes, another is removing the redundant targets, and the third is decomposing the target-coverage graph into independent sub-graphs. It proves that the three approaches are correct, and presents the Construct Independent Sub Graph Algorithm(CISGA) to divide target- coverage graph. Measurement results show that CISGA can reduce 30% of network scale and greatly reduce the algorithm complexity of target coverage problem.
-
Computer Engineering. 2013, 39(1): 116-119. https://doi.org/10.3969/j.issn.1000-3428.2013.01.025In order to ensure the Quality of Service(QoS) in the concurrent multi-path transmission system with the cooperation of multiple networks based on the Wireless Local Area Network(WLAN), a terminal selection method based on the cooperation of WLAN and cellular networks is proposed in this paper. Queuing models are used to analyze the cooperative transmission system. Under the optimal solution of traffic allocation, the optimal slave terminal set is determined by comparing the whole delay of every link corresponding to different terminal set. Simulation results demonstrate that with the optimal slave terminal set, the whole delay of every wireless link can be minimized and the system performance is better.
-
Computer Engineering. 2013, 39(1): 120-124. https://doi.org/10.3969/j.issn.1000-3428.2013.01.026In cognitive radio networks, because of the shadow effect and multi-path fading environment factors, the spectrum sensing performance of single cognitive user is reliable. A double bidirectional reputation based cooperation spectrum sensing framework is proposed. Each second user has their own selfish reputation, which is used to determine whether the user is a malicious user or not, the second user in different regions has different regional reputations, which reflects how the region is affected by shadow effect and multi-path fading effects. The selfish reputations and regional reputations are updated in real time based on the interaction among the users. And according to the different reputation, different weights for cooperative spectrum sensing are given. Simulation result shows that the proposed method can achieve well detection rate in malicious users.
-
Computer Engineering. 2013, 39(1): 125-130. https://doi.org/10.3969/j.issn.1000-3428.2013.01.027According to the P2P streaming media researches in recent years, this paper summarizes the key of P2P streaming media, including the construction of overlay network technology, data scheduling technology and its various optimization techniques. P2P streaming media is a cost-effective and extensible system, and points out that the development direction to further research on the P2P streaming media application and survey, theoretical analysis and modeling, system architecture, combination with video coding and mobile network technology and so on.
-
Computer Engineering. 2013, 39(1): 131-135. https://doi.org/10.3969/j.issn.1000-3428.2013.01.028Aiming at the problem of existing flow statistical sampling in anomaly detection, this paper analyzes the distortion cause that packet sampling and time domain polymerization lead to flow record time series in theory. It uses scale accumulation feature of network anomaly, proposes multi-scale decomposition method based on Non-overlapping Window Aggregation(NOWA), eliminates the signal noise by sampling and time aggregation. Simulation result shows that this method can reduce impact of sampling rate on signal to noise ratio, improve the performance of the anomaly detection.
-
Computer Engineering. 2013, 39(1): 136-139. https://doi.org/10.3969/j.issn.1000-3428.2013.01.029Current watermarking algorithms for vector data are not resistant to geometric transformation attacks. This paper proposes a new blind watermarking algorithm. Douglas-Peucker compression is made to extract feature point. The ratio of the length between feature points is calculated and the ratio sequence is divided by the threshold. According to the watermark information, the feature point is moved to the odd or even interval. Experimental results show that the algorithm is fairly good in imperceptibility and robust to random noise, map cutting, data compression, geometric transformation attack and so on.
-
Computer Engineering. 2013, 39(1): 140-143. https://doi.org/10.3969/j.issn.1000-3428.2013.01.030Most Three-party Authenticated Key Agreement(3PAKA) protocols are susceptible to man-in-the-middle attack. This paper analyzes security of the 3PAKA protocol presented by Tan, and proves that it can not withstand counterfeit attacks and man-in-the-middle attack. To enhance its security, an improved protocol is proposed by one-way hash function and elliptic curve cryptography. Security analysis and formal proof results show that the improved protocol inherits the security of the original protocol, and is able to resist counterfeit attack and man-in-the-middle attack.
-
Computer Engineering. 2013, 39(1): 144-148. https://doi.org/10.3969/j.issn.1000-3428.2013.01.031Security of most message authentication codes depends on Hash function, but some Hash functions have not been proved safe. This paper presents a new kind of message authentication scheme based on point function obfuscator with multi-bit outputs and a Message Authentication Code(MAC) construction scheme including a point function obfuscator with multi-bit outputs and Hash function with key. Analysis results demonstrate security of the scheme depends on the Hash function and point function obfuscator with multi-bit outputs, so it can better guarantee the integrity of the data transmission.
-
Computer Engineering. 2013, 39(1): 149-152. https://doi.org/10.3969/j.issn.1000-3428.2013.01.032This paper proposes a blind watermarking algorithm for double color image based on quaternion. Frequency matirx of the host image and watermark image are gained by using quaternion Fourier transform, and low frequency, middle frequency and high frequency are scrambled by the chaotic sequences. The scrambled watermark is embedded in low frequency, middle frequency and high frequency of host image by using Pulse Code Modulation(PCM). The watermarked host image are gained by using inverse Discrete Quaternion Fourier Transform(DQFT). Experimental results show that the algorithm has the property of perceptual transparency, and is robust to common image attacks, such as cropping image, noise, JPEG compression and so on.
-
Computer Engineering. 2013, 39(1): 153-156. https://doi.org/10.3969/j.issn.1000-3428.2013.01.033Due to the scarceness of resources, most of the proposed grouping-proof protocols are designed using symmetric key cryptographic algorithms, but this may bring security and privacy issues. This paper proposes public-key cryptographic algorithms to satisfy these requirements. In order to overcome the problem of vulnerable to tracking attack of grouping-proof protocol based on Elliptic Curve Cryptography(ECC), an improved scheme is presented to resist tracking attack and analyze the security and performance of the protocol. Analytical results show that the proposed scheme is proved secure against tracking attack, forged proofs attack, impersonation attack and replay attack, and can be applied to Radio Frequency Identification(RFID) system.
-
Computer Engineering. 2013, 39(1): 157-163. https://doi.org/10.3969/j.issn.1000-3428.2013.01.034Aiming at the problem of location privacy leakage in location-based service systems, this paper proposes a method to protect user’s location privacy based on Anonymous Region Transformation(ART). An anonymous region is generated around a location different from the user’s location which the user’s location is not included in, then the neighbor-object-process method is employed to compute the distance between the neighbor objects and the user’s location, so that the accurate inquire results can get while protecting the user’s location privacy. Theory and experimental results show that compared with Cloaking Region(CR) algorithm and SpaceTwist algorithm, the method proposed can guarantee the lower communication overhead, and provide effective location privacy protection.
-
Computer Engineering. 2013, 39(1): 164-167. https://doi.org/10.3969/j.issn.1000-3428.2013.01.035Password-authenticated key exchange protocol can not resist dictionary attack and server compromising attack, so this paper proposes an improved PAKA-X protocol. A user stores his password, while the server stores a verifier for user’s password. Analysis result shows that the proposed protocol has perfect forward secrecy, can resist the Denning-Sacco attack, server compromising attack, on-line dictionary attack, off-line dictionary attack and man-in-the-middle attack. And it has only 9 exponentiation computations, 6 hash-function computations and 6 exclusive-or computations, so it has reliable efficiency.
-
Computer Engineering. 2013, 39(1): 168-174. https://doi.org/10.3969/j.issn.1000-3428.2013.01.036Determining stability problems of Cellular Neural Network(CNN) with delays is researched and a set of sufficient conditions which is used to judge whether cellular neural networks with delays has a unique equilibrium and global asymptotic stability is received. A new Lyapunov-Krasovskii functional is constructed and methods including Lyapunov stability theory and linear matrix inequalities is adopted. The digital simulation results prove effectiveness with feasibility and delay-independent property of the judgment condition. Examples are given to illustrate the good application of the conclusion in secure communication.
-
Computer Engineering. 2013, 39(1): 175-178. https://doi.org/10.3969/j.issn.1000-3428.2013.01.037When the magnitudes of users and commodities grow rapidly, the rating matrix becomes extremely sparse. In the condition, algorithms based on traditional similarity computing have poor performance. In order to overcome this problem, this paper proposes a comprehensive item similarity measurement algorithm based on weighted Jaccard index, and prefills the rating matrix by the comprehensive item similarity. Experimental results show that the algorithm is more accurate compared with traditional algorithms.
-
Computer Engineering. 2013, 39(1): 179-182. https://doi.org/10.3969/j.issn.1000-3428.2013.01.038The original Kernel-based minimum Mean Square Error(KMSE) algorithm, depends on a number of equations to address the multi-class classification problem, which causes a large computational afford. A new KMSE algorithm for Multi-class Kernel-based minimum Mean Square Error(KMSEMC) is proposed for multi-class recognization. It uses only one equation to do so, which is not only mathematically tractable but also computationally efficient. Experimental results on AR face databases show that KMSEMC outperforms KMSE in terms of computation and classification accuracy.
-
Computer Engineering. 2013, 39(1): 183-186. https://doi.org/10.3969/j.issn.1000-3428.2013.01.039In order to improve the classification performance of hypersphere Support Vector Machine(SVM), this paper proposes Fuzzy Minimum Enclosing Ball(FMEB) SVM by integrating several state-of-art classification methods such as Fuzzy Support Vector Machine(FSVM) and Hypersphere Support Vector Machine(HSVM). For pattern classification problem, the basic idea of FMEB is to find two optimal minimum enclosing hyperspheres by introducing fuzzy membership, so that each binary class is enclosed by them respectively, and the margin between one class pattern and the enclosing hypersphere is maximized, thus improving the generalization performance and robustness of hypersphere SVM. Experimental results prove that FMEB is more effective than other methods.
-
Computer Engineering. 2013, 39(1): 187-190. https://doi.org/10.3969/j.issn.1000-3428.2013.01.040In order to reduce color similar background effect on target tracking process, this paper proposes a real-time target tracking method based on object segmentation. Multiple objects are obtained by using background subtraction method. Each target state can be updated by using the position, size and color characteristic of each target. The similar targets in the consecutive frames can be tracked by using matching targets and when the occluded targets separate, the targets can be tracked well by using this feature. Experimental results show that this method is fast and has a desirable tracking result.
-
Computer Engineering. 2013, 39(1): 191-194. https://doi.org/10.3969/j.issn.1000-3428.2013.01.041For improving the learning speed of Support Vector Regression(SVR), this paper proposes a Newton algorithm for Twin Support Vector Regression(TSVR) that tries to find a pair of nonparallel planes by solving two related SVR-type problems and converts the classical Quadratic Programming Problem(QPP) to two small unconstrained optimization problems. Each of the unconstrained optimization problems is solved by Newton algorithm. Experimental results show that the proposed algorithm has good fitting ability as SVR and TSVR, and can reduce the training time.
-
Computer Engineering. 2013, 39(1): 195-199. https://doi.org/10.3969/j.issn.1000-3428.2013.01.042The real-time of existing TV symbol recognition method based on streaming video processing is not good, and has a bad result of recognition to semitransparent one. In order to solve this problem, this paper proposes a recognition method of TV symbol based on improved Chamfer matching. It extracts the TV symbol color and edge gradient direction to represent the shape of it by Canny operator, and uses RANSAC algorithm to optimize and fit the edge. It computes the distance image of edge gradient direction by distance transformation, and identification is based on the principle of shortest distance. Experimental results show that the method can realize the recognition of TV symbol in a single frame and overcomes the difficulty of the recognition of semitransparent TV symbol, while its recognition rate is 97.7% and average recognition time is 801 ms.
-
Computer Engineering. 2013, 39(1): 200-203. https://doi.org/10.3969/j.issn.1000-3428.2013.01.043The appearances of human face, expression, color and so on are different, which bring difficulties to face detection, such as high false face detection and false dismissal. This paper proposes a multi-face detection algorithm based on skin color model and midline localization. Utilizing the facial skin color information firstly, most of the background regions can be excluded quickly. The approximate face candidate regions can be segmented according to facial external structure features. By projecting the edge detection image to X-axes, the midline position of face can be located, and multi-face detection and localization are realized. A black hair single face detection rate of 93.3% can be obtained for a test photo sequence. Experimental results show that the algorithm proposed in the paper can improve the multi-face detection accuracy and is highly robust to facial expression.
-
Computer Engineering. 2013, 39(1): 204-207. https://doi.org/10.3969/j.issn.1000-3428.2013.01.044The traditional nonsmooth Nonnegative Matrix Factorization(nsNMF) is an effective algorithm to deal with hyperspectral data, but the weaknesses of being sensitive to initial values and easily falling to local minimum limit its applying. To solve this problem, the paper proposes a nonsmooth nonnegative matrix factorization algorithm based on Particle Swarm Optimization(PSO) which can find global minimum. It uses the output of nsNMF as the initial values to avoid blind search. It searches endmembers fraction by PSO, then updates endmembers and abundance matrix by nsNMF algorithm. Experimental results based on synthetic data set and truthful data set shows that this algorithm has better global optimal solution. Its endmember and abundance are closer to true value.
-
Computer Engineering. 2013, 39(1): 208-212. https://doi.org/10.3969/j.issn.1000-3428.2013.01.045According to the preparations of oral implant, hard work and low intelligence, an improved Quantum Genetic Algorithm(QGA) is proposed with the information of the dental medical data considered. With the implant localization model studied, this algorithm divides the population to several groups, which respectively takes self-adaptive evolution and communicative crossover operator. Experimental results show that this algorithm can effectively optimize implant positioning parameters, search ability and convergence performance is good compared with the Genetic Algorithm(GA) and Bloch Quantum Genetic Algorithm (BQGA).
-
Computer Engineering. 2013, 39(1): 213-216. https://doi.org/10.3969/j.issn.1000-3428.2013.01.046To help personal sports management system accurately detect users’ activity and give them scientific sports instruction, a common model transferred adaptive activity recognition method is presented. With this method, the common knowledge of many persons can be transferred to a new one and with the new one’s data increasing, the model can be automatically adapted to fit the new user. The method proposed is evaluated on real-world dataset to demonstrate its performance.
-
Computer Engineering. 2013, 39(1): 217-220. https://doi.org/10.3969/j.issn.1000-3428.2013.01.047Existing pixel level edge detection algorithm is dependent on image gray value, and the calculating neighborhood is not clear. Aiming at this problem, an improved method of sub-pixel centre of the circle mark in photogrammetry is proposed. The sub-pixel edge based on digital coarse boundaries using a dual threshold method is found by rotating image coordinate system according to the local gradient direction of the coarse boundaries. An adaptive parameter S in rotated coordinate system is used to search local sub-pixel edge points, and the centre of ellipses is detected on the basis of these points. The proposed algorithm is realized using Matlab, and the experimental results demonstrate its excellent performance and high precision in good agreement with that of the Steger edge algorithm. It is concluded that the new method can satisfy the demands in the opto-electronic measurement.
-
Computer Engineering. 2013, 39(1): 221-224. https://doi.org/10.3969/j.issn.1000-3428.2013.01.048In this paper, an adaptive Shearlet domain regularized minimization total variation image denoising algorithm is proposed, which can overcome the problem of the traditional nonlinear diffusion image denoising methods. The texture and edge domain is adaptively shrinkaged in Shearlet domain. And then, a total variation regularized energy functional model with restrictions on the finite adaptive shearlet domain is used to deal with smooth domain. Experimental results show that this algorithm can reduce the noise and preserve the edge information, especially to the images containing abundant texture.
-
Computer Engineering. 2013, 39(1): 225-229. https://doi.org/10.3969/j.issn.1000-3428.2013.01.049In this paper, an algorithm based on matched filtering and gradient correlation is presented for finger vein certification. A method based on histogram statistics is given, which can distinguish whether the image is finger vein or not. Matched filtering based on maximum curvature model is adopted to capture the gradient image of finger vein. Cross-correlation between two gradient images is figured out to estimate their similarity. The most correlation is threshold treated to decide whether matching or not. Experimental results show that the FAR and the FFR respectively is 0.375% and 1.20%. This algorithm is easy to realize and efficient, insensitive to noise, and its recognition rate is satisfactory, which is more suitable for the small DSP-based finger vein authentication equipments.
-
Computer Engineering. 2013, 39(1): 230-233. https://doi.org/10.3969/j.issn.1000-3428.2013.01.050The encoding time of traditional fractal image coding method is too long. Aiming at this problem, this paper proposes a novel image coding method based on dictionary. The Julia sets can produce abundant images with fewer parameters. Therefore they are utilized to produce the coding dictionary and each image can find the suitable domain block in the dictionary. Experimental results show that the method can improve the coding efficiency significantly in the premise of not reducing the coding quality obviously, especially for the large images.
-
Computer Engineering. 2013, 39(1): 234-238. https://doi.org/10.3969/j.issn.1000-3428.2013.01.051For the problem of using image processing to accurately measure the plant leaf area, this paper proposes a new method which utilizes Snake model to extract the leaf contour and calculates the leaf area based on the leaf contour. It improves the Snake model. On one hand, it defines a color gradient in HSI color space as the external energy function of snake. On the other hand, it makes the corners detected as the vertexes of the initial contour, and increases or reduces the vertexes by adaptive algorithm. It utilizes the Snake model to extract the leaf counter. It calculates the leaf area based on the chain codes of leaf contour. Experimental results prove that, compared with the leaf area calculation method in image J software, this method is more accurate, and it has certain application value.
-
Computer Engineering. 2013, 39(1): 239-243. https://doi.org/10.3969/j.issn.1000-3428.2013.01.052Due to the particularity of the text images, using the Scale Invariant Feature Transform(SIFT) algorithm to extract can not get enough feature points. Otherwise, strong inferences between different words lead the dad matching result. Aiming at this problem, this paper makes a research on the SIFT algorithm and proposes an improved SIFT algorithm for text images template matching. The improved algorithm uses three methods to improve the efficiency, such as to use threshold images instead of gray ones. It increases the number of feature points and cancels the rotational invariance. Experimental result shows that the improved algorithm can effectively improve the accuracy of the text template matching than standard SIFT algorithm.
-
Computer Engineering. 2013, 39(1): 244-247. https://doi.org/10.3969/j.issn.1000-3428.2013.01.053In traditional image median filtering algrithm, two key factors that can effect the denoising power of this filter, namely fixed order window and unconditional median operations. Aiming at this problem, this paper proposes an improved algorithm. The pseudo-quantized representation of digital image is put forward through using quantum theory and an adaptive mechanism is introduced by means of quantum Hadamard transform, so that the size and shape of the order windows can be automatically adjusted according to the local characteristics of the image. Furthermore, a conditional median algorithm is also introduced to preserves image details as much as possible in the denoising. Simulation results show that the proposed algorithm has stronger ability to filter noise and preserve significant image details simultaneously compared with the traditional median filtering algorithm and recursion median filtering algorithm. Another advantage of the proposed improved median filtering algorithm is not sensitive to the strength of the noise, it means that the effection of strength of noise to filtering is very small.
-
Computer Engineering. 2013, 39(1): 248-251. https://doi.org/10.3969/j.issn.1000-3428.2013.01.054Because of noise and low contrast of the infrared image, an infrared image enhancement algorithm is proposed. Evaluating function is constructed by combing information entropy, signal-noise-ratio with standard variance in order to evaluate the enhanced image quality. Thus, an efficient nonlinear transform function is designed by arc-tangent transform in stationary wavelet domain so as to enhance the contrast of an infrared image. The optimal nonlinear transform parameter is obtained by combing differential evolution algorithm with the constructed evaluating function. Experimental results show that the proposed algorithm can efficiently enhance the contrast of the infrared image while getting a good robust to noise in the infrared image.
-
Computer Engineering. 2013, 39(1): 252-255. https://doi.org/10.3969/j.issn.1000-3428.2013.01.055To the problem that objective image fusion quality assessment is not consistent with the subjective quality assessment, this paper puts forward an image fusion quality assessment method based on watershed algorithm. It utilizes the structure similarity theory to acquire the structure similarity information map, and applies watershed algorithm to segment the information map. In order to reduce the excessive segment phenomena, it makes the tag in the watershed and divides the image information into mutual information and independent information. It designs the mapping functions that process the mutual information and independent information separately, and accomplishes the fusion image quality assessment. Experimental results show that, compared with traditional methods, this method has advantage in anti-disturbance.
-
Computer Engineering. 2013, 39(1): 260-263. https://doi.org/10.3969/j.issn.1000-3428.2013.01.057To solve the interoperability difficulty between mobile Agent and Struts 2, combining with the advantages of both, a new Aglet-Struts 2 integrated development framework based on mobile Agent is proposed. This framework integrates mobile Agent with Struts 2 control action, JSP view and javaBean model in together by the agletaction service proxy, and achieves transparent access to database by persistent layer Hiberanate. Using XML technology for data exchange and hierarchical structure for low coupling, maintainability, extensibility and flexibility of the framework are greatly improved. Through the application of the framework to the remote monitoring system of gas production data, the feasibility of the framework is further demonstrated. It provides a new solution and means for complex distributed applications.
-
Computer Engineering. 2013, 39(1): 264-269. https://doi.org/10.3969/j.issn.1000-3428.2013.01.058In order to meet the needs of high-speed multi-channel data transmission and fault-tolerant, this paper puts forward a kind of data transmission management method. Using four-body First Input First Output(FIFO) for data buffering and a channel list for recording the fault state together with a 4×32 matrix switch array, the data transmission between the four-body FIFO and all valid data transmission channel can be organized and managed efficiently. Experimental result in the 8-channel system shows that the maximum communication rate is up to 800 Mb/s, and 7 channels can achieve fault tolerant dynamically.
-
Computer Engineering. 2013, 39(1): 270-274. https://doi.org/10.3969/j.issn.1000-3428.2013.01.059Depending on schedulability problem of the complex real-time system including periodic tasks in multi-processor and multi-partition structure, this paper proposes a tool based on simulation method of schedulability in the task set. Through the setting of clock variables, the tool can simulate task scheduling process, determine the simulation interval according to the characteristics of periodic task set, and optimize the algorithm to determine the schedulability. Test results and analysis verify that this tool has automatic, accurate and fast advantages to describe task scheduling process by Gantt chart. Compared with the other existing tools, it is more accurate and efficient.
-
Computer Engineering. 2013, 39(1): 275-278. https://doi.org/10.3969/j.issn.1000-3428.2013.01.060A model for remote interaction of user with console program is proposed according to the problems that console program compatibility is reduced by adding remote communication module and system application compatibility is reduced by redirecting standard I/O. This model devises a console program loader which adopts a method of writing commands into the console window input buffer by simulated keyboard input and reading echo data from its output buffer for interaction. Experimental results demonstrate that the model is appropriate for all kinds of normal console program and cannot conflict with system software. It can be adopted for interacting with console program.
-
Computer Engineering. 2013, 39(1): 279-282. https://doi.org/10.3969/j.issn.1000-3428.2013.01.061An inappropriate Greatest Available Logical Time(GALT) algorithm may affect system’s whole function and cause deadlock in extreme condition. The deadlock is harmful which prevents system advancing forward. This paper researches deadlock within calculating GLAT, explains reasons of deadlock and proves the reason. It introduces one non-deadlock time management GALT computation algorithm based on Frederick algorithm. In test section, it proves the time advance algorithm’s undoubtedly correctness of non-deadlock.
-
Computer Engineering. 2013, 39(1): 283-286. https://doi.org/10.3969/j.issn.1000-3428.2013.01.062According to the fixed axis rotation linear array camera, this paper puts forward a kind of camera calibration method based on quadratic cuves. It only needs two or more different directions to shooting images, and captures multi-frame image in every direction through the way of camera rotating. A pattern plane including three or more of quadratic is used, and camera and target can move freely, which do not know the motion parameters. Multi-frame linear images shoot in every direction are arranged to area array image by coordinate transformation. The way of using quadratics extracted from area array image as calibration primitive elements simplifies the problem of primitive corresponding. Simulation results show that the method is very robust and has high precision.
-
Computer Engineering. 2013, 39(1): 287-289. https://doi.org/10.3969/j.issn.1000-3428.2013.01.063Aiming at the relatively backward-status of water-quality monitoring for lake and reservoir and the information technology development of water-boom forecasting warning, an intelligent information system including water-quality monitoring, water-boom forecasting and warning is developed. The system is built by the language of C++ of Visual Studio 2010, integrating network communication, Geographic Information System(GIS) and database(SQL2005) key technology. The function of real-time monitor and higher accuracy of long-term prediction for lake water-blooms is achieved through the gray-Back Propagation(BP) neural network model, which provides an effective decision-making platform for environmental protection departments to prevent and control lake water-bloom.
-
Computer Engineering. 2013, 39(1): 290-293. https://doi.org/10.3969/j.issn.1000-3428.2013.01.064Because of the uncertain conflicting information confronted by Multi-source information fusion in radar target recognition, a new conflict evidence integration method based on confidence distance is put forward. This new method takes the relevance of evidence into consideration under the premise of not changing the D-S combination rules, by computing the similarity and credibility of the evidence and using the weighted average of the credibility to replace the conflicting evidence, in order to eliminate the conflict evidence and integrate the information. This method is compared with some existing methods through simulation, which indicates that it is an effective method in conflict evidence integration.
-
Computer Engineering. 2013, 39(1): 294-297. https://doi.org/10.3969/j.issn.1000-3428.2013.01.065Multi-label classification problem is generally converted into multi-class classification problem, and it can increase the amount of computation and extend running time. When a new class is added, the method has a poor expandability too. To solve this problem, this paper proposes a multi-label classification method based on sphere structured Support Vector Machine(SVM). Each category labels correspond a ball domain structure, this paper extracts the samples of ball overlap region, according to the distance difference measures the sample category similarity, and is to determine the category of samples. Experimental results show that the method can save 210 ms training time, improves the average recall rate of 3.2% and is suitable for a large number of sample classification.
-
Computer Engineering. 2013, 39(1): 298-302. https://doi.org/10.3969/j.issn.1000-3428.2013.01.066This paper proposes a multi-view video coding system based on Discrete Wavelet Transform(DWT). The proposed scheme carries out four-dimension wavelet decomposition on multi-view video sequences, including temporal one-dimension decomposition, view one-dimension decomposition, and two-dimension spatial decomposition. Besides utilizing the intrinsic multi-resolution feature of DWT, the proposed scheme incorporates the embedded entropy coder for accomplishing temporal, spatial, view and quality scalabilities. According to various constrain of broad width, the coding stream of each subband can be optimal truncated by rate distortion optimization. Simulation results demonstrate that the reconstructed picture quality is maintained in this coding system.
-
Computer Engineering. 2013, 39(1): 303-308. https://doi.org/10.3969/j.issn.1000-3428.2013.01.067The simulation of control system generally adopts the single processor, but its real-time and efficiency of simulation cannot meet the current needs. The parallel simulation method for the control system is put forward. The segmentation of simulation structure diagram is achieved. According to the results of segmentation, corresponding task graph is established. The task graph has a belt loop, cross-feedback, and characteristic of periodicity. Aiming at this kind of task graph, load balancing dynamic scheduling algorithm is provided. The example proves that the algorithm improves the efficiency of the system, and meets the requirements of real time.
-
Computer Engineering. 2013, 39(1): 309-312. https://doi.org/10.3969/j.issn.1000-3428.2013.01.068To solve the problem that the cnvPartition algorithm runs slowly on the Single Nucleotide Polymorphisms(SNP) data, this paper presents an improved gene Copy Number Variation(CNV) detection algorithm, which is combined the binary segmentation algorithm with the cnvPartition algorithm. This algorithm detects the marked breakpoints first, and then is the breakpoints in small copy number variation regions. Experimental results show that the highest of the detection speed of this algorithm can be increased by 50%, and is easy to detect the vignettes of CNV variation.
-
Computer Engineering. 2013, 39(1): 313-317. https://doi.org/10.3969/j.issn.1000-3428.2013.01.069In order to get the browser IE8.0 login information of Windows 7 operating system, by analyzing two aspects of the login information storage policy and encryption mechanism, in the study of the shortcomings of the protective mechanism. For user’s security settings for IE, this paper proposes a reduction method to intercept the historical records. On the basis of the interception user access the Web site address, it filters out the Web site address to obtain the decryption key, and gets the login information which is automatically saved by users. Experimental results show that this method can effectively obtain the user’s login information which is established by user’s security settings
-
Computer Engineering. 2013, 39(1): 318-320. https://doi.org/miss_hit_529@163.comAs the commonly used software coding method in H.264 standard is difficult to realize the high definition video real time decoding, this paper presents a method of software and hardware of co-design H.264 high definition encoder for software and hardware division and the distribution of the task, and designs its software part and respectively hardware parts. After the completion of the design, build software hardware co-simulation environment for the whole system function and performance simulation, and using the Constrained Random Verification(CRV) method to improve the verification function coverage rate. Simulation results show that the encoder system on 1 024×768 pixel image frames is encoded, fully meets the requirements of real-time encoding.
2024
Volume 50 No 3
Volume 50 No 3
(Started in 1975, Monthly)
Sponsored by：
East-China Institute of Computer Technology Shanghai Computer Society
Edited & Published by：
Editorial Board of Computer Engineering
Chief Editor：Jiang Bo
Executive Chief Editor:
Hu Bing
Tel：021-67092217
E-mail: ecice06@ecict.com.cn
CN: 31-1289/TP
ISSN: 1000-3428
Sponsored by：
East-China Institute of Computer Technology Shanghai Computer Society
Edited & Published by：
Editorial Board of Computer Engineering
Chief Editor：Jiang Bo
Executive Chief Editor:
Hu Bing
Tel：021-67092217
E-mail: ecice06@ecict.com.cn
CN: 31-1289/TP
ISSN: 1000-3428