Search

Copyright of Transactions of Beijing Institute of Technology is the property of Beijing University of Technology and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Transactions of Beijing Institute of Technology is the property of Beijing University of Technology and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

In modern electronic warfare, the GNSS repeater deception jamming has been an important method of jamming. With the research on satellite selection algorithm of GNSS repeater deception jamming, the jamming ability of repeater deception jamming can be improved. A satellite selection algorithm was proposed based on the satellite selection method of common GNSS positioning and PDOP contribution of single satellite. The simulations were carried out based on the pseudorange single positioning method with the least-squares algorithm, under the static and dynamic conditions respectively. And the receiver attacked by repeater deception jamming with single satellite was simulated under both static and dynamic conditions. The simulation results show that the satellite selection algorithm for repeater deception jamming proposed in this paper can provide the theoretically optimal star selection result with high probability. [ABSTRACT FROM AUTHOR]

Copyright of Transactions of Beijing Institute of Technology is the property of Beijing University of Technology and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Transactions of Beijing Institute of Technology is the property of Beijing University of Technology and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Transactions of Beijing Institute of Technology is the property of Beijing University of Technology and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Transactions of Beijing Institute of Technology is the property of Beijing University of Technology and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

In this paper, an integrated fault diagnosis and fault tolerant control algorithm was proposed for a non-Gaussian nonhnear stochastic distrihution control system. The RBF neural networks observer based fault diagnosis and PI tracking fault tolerant control were integrated to he designed. The rational square-root B-spline model was used to represent the output probability density function (PDF), On the basis of designed nonlinear neural network observer, a new fault diagnosis algorithm was developed to diagnose the slow-varying fault in the dynamic part of such systems, Convergency analysis was performed for the error dynamics raised from the fault detection and diagnosis phase. With the information of fault diagnosis, a new fault tolerant control scheme based on PI tracking strategy was designed so that the post-fault probability density function could still track the given distribution, A simulated example has been given to illustrate the efficiency of the proposed algorithms. [ABSTRACT FROM AUTHOR]

A predictive guidance scheme based on B-plane parameters is proposed for interplanetary low-thrust trajectories in this paper. Firstly, thrust control parameters are piecewise parameterized. The sequential quadratic programming algorithm is used to obtain the optimal nominal trajectory. Then, the coefficients of piecewise function are selected as the guidance parameters, and B-plane parameters are selected as the terminal constraints, so that the function relation between B-plane and guidance parameters is established. Based on differential correction algorithm, piecewise successive correction strategy is used to correct the guidance parameters. Finally, Monte Carlo analyses of Earth-Mars transfer are used to verify the efficiency of the proposed method. The result indicates that, compared with the orbit parameters method, the B-plane parameters method has the higher guidance precision and less correction frequency. [ABSTRACT FROM AUTHOR]

A segmentation algorithm for adherent cell images based on edge chain-code information is presented in this paper. Firstly, a pre-processing was carried out for the cell image with low contrast ratio. Then edge chain code of cell binary image was used to keep track of their contours and extract their shape features, such as chain code sum, chain code diff, curvature radius and approximate perimeter. Further, the criterion of distinguishing edge inflexion and sleek curve section was provided. In the end, the adherent cell images were separated through linear interpolation between authentic corners. The presented algorithm has been applied to two sequence images with 120 frames. The segmentation results show that proposed algorithm can not only solve the problem of segmentation but also repair cell edge hollow as well as can count the number of cells successfully. Compared with the threshold algorithm and priori model algorithm, this algorithm can improve the success rate by 40% ~ 60%. [ABSTRACT FROM AUTHOR]

Voice over IP (VOIP) is becoming popular over IP network. However, the VOIP performance would become unsatisfactory particularly when the link is congested. This paper describes a method of improving the quality of service (QoS) of VOIP with low complexity by intensively studying the effects of random early detection (RED) on VOIP. It is found that RED is able to reduce the delay distribution, jitter, effective loss and consecutive loss of VOIP traffic, thus the quality of service of VOIP is improved. In addition, the application of RED algorithm to VOIP could control the traffic congestion to a certain extent. [ABSTRACT FROM AUTHOR]

A novel method of mass segmentation in mammogram is proposed in this paper. First, a mathematical model is presented based on the clinical features of the mass and a multi-filtering method is used to detect the mass' location. Then, according to the result of clustered pixels, which is got by mean shift algorithm, rough handling masses could be obtained. Finally, the active contour without edge model is applied to refine the rough-wrought masses. Experiments implemented on the public MIAS database indicate that the proposed method can achieve better true positive rate than the morphological component analysis method with low false positive rate. In addition, the proposed segmentation method could emerge complete boundary, which could meet the clinical examination and diagnosis demand. [ABSTRACT FROM AUTHOR]

To deal with the shortage of domain ontology to describe semantic based fuzzy knowledge, this paper presents a four-layered extended structure of the fuzzy ontology by using the advantages of fuzzy theory to deal with fuzzy information, so as to extend existing domain ontology. Fuzzy concepts, fuzzy properties and fuzzy memberships in fuzzy ontology are defined to realize the description of fuzzy knowledge of ontology by semantics. Using the four-layered extended structure, an expert fuzzy ontology in the domain of science and technology evaluation was built, and the expert selection algorithm based on fuzzy ontology was designed to realize expert selection service according to the fuzzy semantic information in selection requirements. The practical application results show that semantic expert selection service is able to select the appropriate experts efficiently and accurately. [ABSTRACT FROM AUTHOR]

To track pedestrian oriented multiple objects in video sequence, this paper presents a tracking algorithm using multi-cue integration approach and high degree occlusion processing. Combined with the mean-shift algorithm, multi-cue integration method could track object in normal condition by fusing the color and motion information. To deal with the issue of occlusion involving multiple objects, the area change of the object during occlusion has been analyzed in theory. An occlusion factor is advanced to detect occlusion occur, identify obstructer or occluded object and to check the reappearance of an occluded object. Experiment results show that the proposed tracking algorithm could correctly find out and resolve the occlusion among the multiple objects so that it could successfully track the multiple pedestrians. [ABSTRACT FROM AUTHOR]

The application of quantitative feedback theory (QFT) to design the flight control system combined with active disturbance rejection control (ADRC) algorithm for some missile is studied in this paper. From the results of simulation, the robustness analysis and the comparison of the ADRC and the PID algorithm, it could be concluded that the designed control system has good responses and robustness with satisfactory control performance. The output response overshoot of ADRC is less than PID control, but the rapidity is slightly worse than PID control. The ADRC algorithm to design the height keep loop of the missile longitudinal control system can get good tracking performance. The vertical speed tracking has hardly any overshoot, which shows good dynamic performance and steady state performance. [ABSTRACT FROM AUTHOR]

Existing routing algorithm for triplet-based hierarchical interconnection network (THIN) could not be applied to non-balanced THIN. To solve this problem, this paper proposes a new shortest path routing algorithm (SPORT). It computes the shortest path at the source node and constructs the output port lists, which can apply to both balanced THIN and non- balanced THIN. The platform of experiment was established by utilizing Noxim simulator. Simulation results show that the traffic latency of SPORT is lower than existing routing algorithms such as DDRA and Min-DDRA ones. Furthermore, when the localization is considered to compare the latency of THIN with 2D-mesh, results show that, for those networks with high-localized workloads, the performance of THIN is superior to that of 2D-mesh. [ABSTRACT FROM AUTHOR]

A new connected components labeling algorithm is proposed in this paper. Compared with other algorithms suitable for hardware, it not only increases the function of false alarm removing, but also compounds a one-dimensional array and run-length codes structure into a two-dimensional array structure. The proposed algorithm has great advantages in terms of speed, resource utilization and algorithm performance. Results of simulation based on Matlab and FPGA demonstrate that new algorithm can completely record labels and parameters information. Compared with other algorithms, time consumption would decrease approximately fifty percent and the complexity of the controlling logic is reduced, when this algorithm is used to meet the requests of FPGA application to acquiring parameters' information with high speed and preciseness. [ABSTRACT FROM AUTHOR]

To improve the performance of general bearing estimation algorithm using acoustic vector sensor array (AVSA) with rotational invariance technique (ESPRIT), an improved algorithm is proposed and carried out in this paper, which is based on the combination processing of pressure and particle velocity. The cross-covariance matrix of pressure and particle velocity was constructed by taking the mean value of cross-covariance matrixes under even-spacing driving angles and using ESPRIT. Simulation and test result on the lake environment show that the proposed algorithm is more effective than existing algorithms in resolution and accuracy under the case of low signal-to-noise ratio. The foundation o{ improved algorithm is the principle of coherency between pressure and particle velocity, which could suppress interference strongly in isotropic noise field. Moreover, with the use of ESPRIT algorithm, the AVSA could obtain much more gain and have advanced performance in azimuth angle estimation. [ABSTRACT FROM AUTHOR]

This paper presents a novel particle-based pedestrian tracking algorithm for PTZ camera surveillance. Most of the state-of-art particle-based tracking algorithms are challenged due to lacking of a reliable moving object detection and drastic scale along with perspective shift of the target. Therefore, pure intensity based algorithms usually miss the target gradually without other features for correcting target location. Our method learns and maintains a contour template of the target besides intensity. Taking into account both the evolution and sudden change of the pedestrian contour, the proposed tracking algorithm maintains several sets of profiles from different perspectives and evolves them incrementally. The effectiveness of our tracking algorithm with extra contour measurement has been tested over several surveillance records captured from PTZ camera. Compared with other cutting edge tracking algorithms, this presented algorithm could estimate the target location more robustly. [ABSTRACT FROM AUTHOR]

To improve the retrieval speed of the map in an embedded navigation system, an algorithm of dynamic partition is introduced in this paper. According to the number of map features, the algorithm divided the map into several grids. Through increasing the store space cost to reduce the retrieval time, the real time performance of the algorithm was improved. After map partition, the fusing map matching algorithm was realized, which amalgamated road topological relation and D-S evidence reasoning. The simulation result shows that the proposed algorithm has better matching accuracy and it could improve the real time capability and adaptability of the navigation system. [ABSTRACT FROM AUTHOR]

This paper concerns the time and frequency domain analysis of the along-track motion error in synthetic aperture radar (SAR). The signal model with motion error is established. The effects on the range migration and azimuth modulation are obtained. Based on the series reversion laws, the two dimensional point target spectrum is derived. Then, the error features are formulated. Analysis and simulation results show that the echo characteristics in frequency domain are more sensitive to the motion error. Above-mentioned study provides a theoretical basis for system analysis and imaging algorithm design of SAR. [ABSTRACT FROM AUTHOR]

Because of the atmospheric disturbance and the detector noise, lidar echo signal usually contains a lot of random noise. It is important to smooth the lidar signal before retrieving the atmospheric parameters. A iteration method base on half-step interpolation (HSI) is introduced in the paper. Using the raw data as the original signals, the proposed method could reconstruct the new data by repeating the HSI process until the similarity of adjacent iterated results being smaller than a setting criterion. One of the advantages of the new method is that the data segmentation processing could be omitted. Experiments show that the algorithm can effectively improve the smoothness of the lidar echo and can save the important signal fluctuation adaptively. [ABSTRACT FROM AUTHOR]

Focusing on the ambiguity problem of frequency offset estimation in DSSS signal synchronization, this paper presents a double acquisition algorithm of large frequency offset. This algorithm includes three steps. Firstly, pre-estimation and compensation of frequency offset is implemented by searching time-frequency domain in parallel. Secondly, frequency estimation points are constructed according to the result of pre-estimation and compensation. Thirdly, the frequency ambiguous of the signal is eliminated by double acquisition after frequency compensation. Finally, the correct signal can be obtained after the process in large frequency offset environment. Simulation and real test results show that the acquisition algorithm, which combines the techniques of large frequency offset pre-estimation based on the time-frequency search and fine frequency offset estimation based on FFT and double capture, can effectively shorten the time of DSSS acquisition and improve the acquisition probability in low SNR. The algorithm can be employed in DSSS receiver to satisfy the requirements of simple, flexible and high accuracy. [ABSTRACT FROM AUTHOR]

In order to recognize the criminal suspects in the crowd relevant to the case by computer and improve the efficiency of solving the case, this paper presents a new method of suspected culprit recognition based on Probit model. Separation algorithm, correlation algorithm based on cluster and the parameters named significance level in Probit model are adopted to find the important attributes of the culprit, and then, by training the data just including the important attributes, the crime risk evaluation model could be achieved. The experimental results show that average accuracy of model-recognizing-culprit is 90.5% and the average recall is 92.7%. [ABSTRACT FROM AUTHOR]

By investigating the working process of tanker trucks in a military airport, this paper proposes a centralized channel assignment algorithm and a power rating control method for fast moving nodes to improve the real-time property and energy efficiency of the information collection, detect the state of mobile devices promptly, reduce interference between wireless networks and to increase network capacity in oil security systems. Considering the fact that IEEE 802. 11 wireless LAN standard defines a number of non-overlapping channels which could be used at the same time, two wireless network interfaces operating on different channels could be installed to every node. Through the negotiation on the control channel and switching the data channel dynamically, the balance of network loads could be fulfilled so as to increase the total bandwidth. Simulation and application results demonstrate that, only three data channels is needed to realize the centralized channel assignment and the power rating control, and the network capacity and energy efficiency could be improved effectively. [ABSTRACT FROM AUTHOR]

Based on the derivative constraints, a reduced-rank cross-spectral metric (CSM) adaptive beamforming algorithm is proposed in this paper. In the proposed algorithm, the derivative constraints are firstly applied to impose a constrained band on the direction of expectation, and then the reduced-rank data is generated by using the reduced-rank adaptive signal processing based on CSM to generate the adaptive beamforming. The proposed algorithm is not only robust to the steering vector mismatch between the presumed and the actual array steering vectors, but also reduces the computational load significantly relative to the conventional robust adaptive beamforming algorithms. Simulation founds that the proposed algorithm can improve the performance significantly. [ABSTRACT FROM AUTHOR]

To deal with the issue of error ambiguity-resolving under high phase noisy circumstance in multi-baseline phase interferometer, this paper presents an improved algorithm of successive recursive ambiguity-resolving based on the vector average filtering of phase-difference. The proposed method calculates the vector average values of the multi-point phase-difference. The impact of phase noise on ambiguity-resolving is reduced by averaging vectors between channels so as to improve the accuracy of phase ambiguity estimation at all levels. The correct ambiguity resolving could be achieved even in higher phase noise conditions. The computer simulation results demonstrate the merits of presented algorithm [ABSTRACT FROM AUTHOR]

A new image edge detection algorithm which is based on the theory of fuzzy gradient is proposed in the paper. The steps of the algorithm is as follows: firstly, the original gray image is changed into fuzzy matrix by the normalized tangent function; secondly, gradient matrix of the fuzzy matrix mentioned above is computed; thirdly, the fuzzy gradient matrix which is also called fuzzy set of the gradient matrix is obtained using sine function; and in the end, appropriate threshold which is identified by genetic algorithm based on normalized real-coded is used to segment the fuzzy gradient set into two subsets: the edge point set and non-edge point set. At this point, the image edge points are obtained. To test the speed and accuracy of the algorithm for detecting edge of micro-part image, the experiments have been done taking different kinds of micro-gear as the subjects. The experimental results show that the detecting accuracy is better than 10µs and the processing time is generally about 1 s. Compared with Canny or Pal-King algorithms, it could be concluded that the performance of proposed new algorithm is the best. [ABSTRACT FROM AUTHOR]

Key pre-distribution and management scheme of wireless sensor network based on Hash constructed by look-up table is proposed in this paper. The scheme dynamically generates the public keys for nodes communication through the look-up table encryption algorithm, which can encrypt and hash the text at the same time. To improve the connectivity and livability of the network and also to save the memory, the clustered sensor network style is introduced. Theoretical analysis and computer simulation indicate that the presented scheme can improve the performance of security, connectivity, anti-captured ability and can effectively utilize the memory as well. [ABSTRACT FROM AUTHOR]

This paper presents a watermarking algorithm based on non-negative matrix factorization (NMF) and orthogonal projection transformation. Firstly, the base matrix of an image is constructed by NMF, which is the part-based image representation. Then, the base matrix is converted perpendicularly and it is regarded as a key to extract the watermark later. The watermark is embedded into the coefficient matrix, which is made by mapping the image to the orthogonal base matrix. Finally the image is reconstructed by inverse transform of the matrix. The presented algorithm demonstrates an obvious advantage in precision of reconstruction because the basic NMF method is improved by above steps. It has been applied to the digital watermarking system and the result shows better robustness and practicability than that reported in reference [4]. [ABSTRACT FROM AUTHOR]

In this paper, we propose a new rapid recovery method for continuous data protection (CDP) system to solve the problem in traditional CDP system that it takes large amount of time to recover to any point in time. The proposed method could be called by the technique of back-link based recovery on demand. After discussing the data structure of new method and introducing the concept of back-link based on the pointers in the data structure, we have developed and constructed the new rapid recovery algorithms. The test has been implemented by using typical workload trace data. Experimental result shows that, when the recovery point is near the current time point, the rapid recovery method can present the temporary disk image with little loss of response time. [ABSTRACT FROM AUTHOR]

This paper proposes a routing algorithm by using the ant colony optimization that is applicable to WMSN. The algorithm simulates the means of ant colony finding their ways in nature. One node calculates the probability to choose next hop to send package on the basis of the frequency used prier to this route and the cost through this path. The experimental results show that, compared with traditional direct diffusion algorithm, the average service life lengthens by 20%, and the speed of successfully sending packages is much higher by using the proposed algorithm. [ABSTRACT FROM AUTHOR]

Object recognition with changes in 3D viewpoint is a hot spot currently in the domain of image pattern recognition. An algorithm for pattern matching based on Hausdorff distance is proposed in this paper. The principal components transformation is introduced to this algorithm and makes it stable, distinct and almost real time. Meanwhile, a method for extracting corner feature from images is proposed. The extracted feature is invariant to translation, rotation, scale change and is shown robust to addition of noise. Further, a system to recognize the object with changes in 3D viewpoint is presented using this feature and BP network. The results of experiments for comparison demonstrate that the proposed method is more effective than the other three ones. [ABSTRACT FROM AUTHOR]

This paper presents a revised type of Asymmetric AdaBoost algorithm--M-Asy AdaBoost. It can ensure the success of training process by sample weight distribution. The weight of classifier is optimized by adopting error rates of positive sample. The capability of recognizing positive sample is enhanced. The detection probability is improved and increased monotonously by restricting the classifier adding in the ensemble. The proposed algorithm can attain high detection rate with low false alarm ratio. In the end, it is proved by computer simulation that the M-Asy AdaBoost is effective. [ABSTRACT FROM AUTHOR]

A path planning algorithm based on convex hull is applied to autonomous service robot in this paper. Firstly, the coordinates of ball positions are determined by using Haar-like features classifier, the coordinate of robot's position and its heading are determined by using the method based on color model. According to the features of the robot itself, the balls in a certain range can be treated as one target point. Then a better ball-picking path is elaborated with a path planning algorithm based on convex hull, which takes the target points as the input of algorithm. The algorithm alleviates the robot's motion cost and improves the ball-picking efficiency of the autonomous service robot obviously. [ABSTRACT FROM AUTHOR]

To enhance the performance of speech coding strategy of cochlear implants, wavelet transform zero-crossings stimulation (WZCS) strategy based on one-octave wavelet transform is presented. Zero-crossings of processed signal were derived in the domain of wavelet transform and further the amplitude of signals was modulated by corresponding temporal envelope to form the synthesized stimulus. Four strategies of WZCS, continuous interleaved sampling (CIS), channel specific sampling sequences (CSSS) , frequency amplitude modulation encoding (FAME) were simulated using Matlab. Correlativity analysis between synthesized stimuli and original signal, also between their spectra and that of original speech, was implemented. Compared to other 3 strategies, WZCS obtains the highest correlation coefficient between spectrum of synthesized signal and that of original speech, that is to say, the spectrum of synthesized signal through WZCS strategy is the most correlated to that of original speech. The algorithm presented in this paper keeps more frequency components and phase information of original speech in synthesized stimulus compared to other strategies. [ABSTRACT FROM AUTHOR]

Vision guidance is one of the key techniques for autonomous robot navigation, which allows a robot to find a valid path and recognize the environment. This paper analyses and investigates the problems of image-based road detection and understanding. Based on the colour features of the road, an improved region-growing algorithm is used to segment the images. Due to the influence of environment illumination and disturbances, some lane regions may be lost by over segmentation in the image. To improve the accuracy of lane segmentation, wavelet-based texture features and space adjacencies are employed to retrieve the lost lane regions. The experimental results demonstrate that the proposed method can achieve full-image segmentation and is of high precision, robust and reliability for real-time road segmentation and detection. [ABSTRACT FROM AUTHOR]

In this paper, a novel algorithm, combining trust region and particle filtering, is proposed for target tracking in FLIR imagery. Trust regions and particle filters are two successful methods for object tracking. Both of them have their respective advantages and disadvantages. The proposed algorithm integrates the advantages of the two approaches. Compared with the original particle filters and trust regions, this new algorithm can maintain multiple hypotheses with fewer particles by encouraging the particles to be in the right part of the state space. Thus, the sample degeneracy problem and sample impoverishment problem could be overcome. The experiments performed on several infrared sequences show the robustness of the proposed algorithm. [ABSTRACT FROM AUTHOR]

Combining with the practical application of driving environment, this paper proposed a vehicle detection method based on improved DBSCAN clustering algorithm using a laser scanner. First, the vehicle and sensor coordinate conversion model was built to fuse the data from the laser scanner and the camera. Then the DBSCAN algorithm was improved to cluster the laser scanner data points and remove the noises at the same time. Based on the models of vehicle shape features, the preceding vehicles could be detected using shape matching. Finally, the result of the detection was projected onto the video image. The tests on running vehicle show that the proposed method can detect the vehicles efficiently in real traffic environment. [ABSTRACT FROM AUTHOR]

Background prediction is one of the main background suppression methods for target detection. A novel algorithm of background prediction based on the chaotic character of the clutter cloud is presented in this paper. It is realized that the cloud background of IR image has the chaos character in space, and the Volterra adaptive filter can be introduced to forecast the clutter background. Simulation results show that the proposed algorithm has good effect on cloud background suppression and the residues have white noise characteristics. It also improves the forecast accuracy of the clouds background and obviously increases the signal-to-clutter ratio (SCR). [ABSTRACT FROM AUTHOR]

It is known that particle filter has the superior characteristics of solving non-linear non-Gaussian problems. A new regularized particle filter is proposed to improve the quality of image restoration with mixed or multiplicative noise in this paper. To reduce the error of sample importance re-sampling (SIR) particle filter, which comes from the neglect of measuring data when re-sampling, the posterior continuous distribution sample is adopted. The combination of the cumulative distribution function (CDF) and regularized re-sampling step makes the algorithm have the advantages of minimized variance, alleviating degradation and escaping from the exhaustion of particle. The proposed method has been applied to the mixed noisy and medical multiplicative noisy image restoration. The results show the effectiveness of the algorithm, and demonstrate its superiority, compared with wavelet threshold shrink method and SIR particle filter method. [ABSTRACT FROM AUTHOR]

A fast set partitioning in hierarchical tree (SPIHT) algorithm is proposed in this paper. Compared to the typical SPIHT algorithm, three measures have been taken to speed up SPIHT algorithm. First, human visual system (HVS) is applied to weight the wavelet coefficients, and thus the impacts of subjective fators on image coding are paid more attention. Second, a list of maximum pixels is calculated and introduced so as to, in the process of scanning coefficients, only the corresponding data in such list is necessary to be compared with the threshold. Third, more than one spatial orientation trees are parallel encoded. Experimental results show that the modified SPIHT has double implementation efficiency in terms of execution time, while the reconstructed image has better visual quality with an average gain of 0.2 dB PSNR. [ABSTRACT FROM AUTHOR]

In this paper, a human face tracking algorithm combing color and texture features based on particle filter is proposed. The proposed approach makes use of the characteristics of particle filter which not only can effectively deal with nonlinear and non Gaussian process but also can combine multiple face features information. Different robustness to different environments of features is considered and the weighted color histogram and rotated complex wavelet filter (RCWF) are used to describe features. Thus the color and texture features can be fused under particle filter framework to develop the new face tracking algorithm. Experimental results demonstrate that, compared with the method based on single feature, the robustness, accuracy and flexibility of the algorithm have the better effectiveness of the performance, and the experiments also show the accuracy for face tracking in real scenes. [ABSTRACT FROM AUTHOR]

To show the orientation and magnitude of velocity in flow field, a new 2D flow field visualization method is proposed in this paper by combining the line integral convolution (LIC) algorithm with the HSV color model. With the HSV model, the orientation of velocity is mapped to hue, the magnitude of velocity is mapped to saturation, and the LIC value from LIC algorithm is associated with the value of HSV model. After these processing, a flow field can be visualized as a new color image. This image has not only the same high resolution quality as LIC image, but also shows the orientation and magnitude of flow velocity. The proposed method has been successfully applied to the visualization of 2D explosion flow field. Experiment results show that it can improve the flow visualization effectively. [ABSTRACT FROM AUTHOR]

In documents classification, utilization of ontology and context is proven an effective way to improve text classification, but also presented as a difficult problem. This paper presents a novel adaptive hierarchical classification model (HAC) which is based on concept expansion with ontology. Using the hierarchy structure of the model, an incremental learning algorithm in terms of both feature and category associated context is presented, and then with these, a more efficient and blocking-immuned hierarchical classification method is proposed. Experimental results showed that better performance can be achieved with less classification time in HAC, which is of particular significance in document online classification. [ABSTRACT FROM AUTHOR]

Pattern search is the essential part of all fast search motion estimation algorithms of MPEG-4 or H. 264 standard. Efficient search patterns can improve the performance of the existing fast search algorithm. In this paper is analyzed the computation distribution of the block-based fast search motion estimation algorithm and the performance of different search patterns. In order to speed up the motion search procedure while maintaining the quality of the decoded sequence, we chose the double search patterns strategy, and suggested an adaptive pattern selection method. The experimental result showed that the small pattern can improve the search efficiency and the large pattern can guarantee the search accuracy. This adaptive double pattern search algorithm can achieve very similar PSNR and bits performance compared with partial distortion fast full search algorithm, while it has a speedup of 151 in real-time. [ABSTRACT FROM AUTHOR]

The design of a non-magnetic meter chip is introduced in this paper. It is based on the sensor circuit design, reflecting a high-sensitivity algorithm in its system architecture. Via the I²C interface, it can be intelligently controlled to adjust its operating mode, sensors setting, sampling frequency et al. This meter chip is already taped-out in 0.6 µm process and tested, the results showed that the chip can measure 90 degree turn of the impeller. The current is 5 µA in the working mode and only 0.01 µA in the sleep mode with 3.3 V power supply. The chip can be widely used in flux and heat meters. [ABSTRACT FROM AUTHOR]

In order to study the physical mechanism of explosion field in concrete, a code MMIC-3D is proposed based on the 3D Youngs' interface reconstruction algorithm. The formation and development process of the cavity and the law of bulge motion during the explosion in semi-infinite concrete are simulated. The three-dimensional numerical results of the explosion field in the semi-infinite concrete medium are obtained. The results showed that the computational data agree with the experimental phenomena and the physical law, which indicates that the algorithm and the hydrocode presented in this paper are correct, and can provide instruction for engineering designs. [ABSTRACT FROM AUTHOR]

By stochastic analysis method and optimal control theory, the paper discusses a class of singular stochastic control problems with stopping time, introduces a drift parameter into the state and extends the diffusion parameter from 1 to σ. The solution of the variational inequation, together with the optimal strategies are given, i.e. the optimal control and the optimal stopping time. The algorithm also shows that solution is just the optimal cost function. [ABSTRACT FROM AUTHOR]

In this paper, a new algorithm based on graph cuts to solve stereo matching problem is proposed. In the algorithm, the constructed graph for minimizing the energy function does not use any auxiliary nodes, therefore smaller graph can be constructed efficiently and the minimum cut can be computed faster. Experimental results demonstrated that satisfactory results can be computed efficiently by the proposed method. [ABSTRACT FROM AUTHOR]