Adaptive learning approach to landmine detection

We consider landmine detection using forward-looking ground penetrating radar (FLGPR). The two main challenging tasks include extracting intricate structures of target signals and adapting a classifier to the surrounding environment through learning. Through the time-frequency (TF) analysis, we find that the most discriminant information is TF localized. This observation motivates us to use the over-complete wavelet packet transform (WPT) to sparsely represent signals with the discriminant information encoded into several bases. Then the sequential floating forward selection method is used to extract these components and thereby a neural network (NNW) classifier is designed. To further improve the classification performance and deal with the problem of detecting mines in an unconstraint environment, the AdaBoost algorithm is used. We integrate the feature selection process into the original AdaBoost algorithm. In each iteration, AdaBoost identifies the hard-to-learn examples and a new set of features which provide the specific discriminant information for these hard samples is extracted adaptively and a new classifier is trained. Experimental results based on measured data are presented, showing that a significant improvement on the classification performance can be achieved.     

Blind adaptive decision fusion for distributed detection

We consider the problem of decision fusion in a distributed detection system. In this system, each detector makes a binary decision based on its own observation, and then communicates its binary decision to a fusion center. The objective of the fusion center is to optimally fuse the local decisions in order to minimize the final error probability. To implement such an optimal fusion center, the performance parameters of each detector (i.e., its probabilities of false alarm and missed detection) as well as the a priori probabilities of the hypotheses must be known. However, in practical applications these statistics may be unknown or may vary with time. We develop a recursive algorithm that approximates these unknown values on-line. We then use these approximations to adapt the fusion center. Our algorithm is based on an explicit analytic relation between the unknown probabilities and the joint probabilities of the local decisions. Under the assumption that the local observations are conditionally independent, the estimates given by our algorithm are shown to be asymptotically unbiased and converge to their true values at the rate of O(1/k/sup 1/2/) in the rms error sense, where k is the number of iterations. Simulation results indicate that our algorithm is substantially more reliable than two existing (asymptotically biased) algorithms, and performs at least as well as those algorithms when they work.     

图像融合技术检测低可观测目标评述

阐明了利用图像信息空间覆盖范围大的特点,运用图像融合的手段来对低可观测目标进行检测.通过综合处理来自多个传感器图像包含的检测对象信息、环境特性信息、运动信息、空间信息、时间信息,获得的融合信息包含任何单一传感器无法提供的信息,进而提高检测性能.扼要介绍了像素层图像融合、特征层图像融合以及符号层图像融合的基本概念.列举了图像融合技术一些应用实例,阐述了用图像融合技术检测低可观测目标的优点.着重强调了检测低可观测目标需要进一步研究的问题.     

Monitoring and Voting in Asynchronous Sampling

A method for treating monitoring and voting for quadredundant, asynchronous signals is described. Each pair of signals is compared to determine whether or not they are in tolerance, and the signal state and fault pattern are determined based on the results of the comparisons. Keys of movement between the fault patterns are derived and used to devise monitoring algorithms for determining the set of redundant signals that should be used by the votor.     

Hypothesis management for information fusion

The efficient management of large collections of fusion hypotheses presents a critical challenge for scaling high-level information fusion systems to solve large problems. We motivate this challenge in the context of two Alphatech research projects, and discuss several partial solutions. A recurring theme is the exploitation of space-efficient, factored representations of multiple hypotheses to enable an efficient search for good hypotheses.     

Sensor fusion for intelligent alarm analysis

The purpose of an intelligent alarm analysis system is to provide complete and manageable information to a central alarm station operator by applying alarm processing and fusion techniques to sensor information. This paper discusses the sensor fusion approach taken to perform intelligent alarm analysis for the Advanced Exterior Sensor (AES). The AES is an intrusion detection and assessment system designed for wide-area coverage, quick deployment, low false/nuisance alarm operation, and immediate visual assessment. It combines three sensor technologies (visible, infrared, and millimeter wave radar) collocated on a compact and portable remote sensor module. The remote sensor module rotates at a rate of 1 revolution per second to detect and track motion and provide assessment in a continuous 360° field-of-regard. Sensor fusion techniques are used to correlate and integrate the track data from these three sensors into a single track for operator observation. Additional inputs to the fusion process include environmental data, knowledge of sensor performance under certain weather conditions, sensor priority, and recent operator feedback. A confidence value is assigned to the track as a result of the fusion process. This helps to reduce nuisance alarms and to increase operator confidence in the system while reducing the workload of the operator     

The case for like-sensor predetection fusion

There has been a great deal of theoretical study into decentralized detection networks composed of similar (often identical), independent sensors, and this has produced a number of satisfying theoretical results. At this point it is perhaps worth asking whether or not there is a great deal of point to such study-certainly two sensors can provide twice the illumination of one, but what does this really translate to in terms of performance? We take as our metric the ground area covered with a specified Neyman-Pearson detection performance. To be fair, the comparison will be of a multisensor network to a single-sensor system where both have the same aggregate transmitter power. The situations examined are by no means exhaustive but are, we believe, representative. Is there a case? The answer, as might be expected, is “sometimes.” When the statistical situation is well behaved there is very little benefit to a fused system; however, when the environment is hostile the gains can be significant. We see, depending on the situation, gains from colocation, gains from separation, optimal gains from operation at a “fusion range,” and sometimes no gains at all     

Design-space adaptation method for multiobjective and multidisciplinary optimization

This paper developed a new method that adaptively adjusts a design space by considering the actual solution distribution of a problem to overcome the conventional design-space adaptation method that assumes the solutions distribution to be a normal distribution because the distributions of solutions are rarely normal distributions for real-world problems.The developed method was applied to nineteen multiobjective test functions that are widely used to evaluate the characteristics and performance of optimization approaches.The results showed that this method adapted the design space to an appropriate design space where the solution existence probability was high.The optimization performance achieved using the developed method was higher than that of the conventional methods.Furthermore,the developed method was applied to the conceptual design of an unmanned spacecraft to confirm its validity in real-world design and multidisciplinary-optimization problems.The results showed that the Pareto solutions of the developed method were superior to those of conventional methods.Additionally,the optimization efficiency with the devel-oped method was improved by more than 1.4 times over that of the conventional methods.In this regard,the developed method has the potential to be applied to complicated real-world optimiza-tion problems to achieve better performance and efficiency.     

An efficient algorithm for multisensor track fusion

An efficient algorithm for track-to-track fusion by incorporating cross-covariance between tracks created by dissimilar sensors is described. An analytical solution of this problem is complicated if cross-correlation between sensors tracking the same target is taken into account. An explicit solution of the cross-covariance matrix at steady state is derived in terms of an integral. It is shown that solution of this integral involves inversion of a matrix whose elements are functions of parameters of individual trackers. Structure of this matrix is analyzed. An efficient analytical solution for inversion of this matrix is obtained. For fusion of similar sensors, it is shown that this matrix is reduced to the Routh-Hurwitz matrix which arises in the study of steady state stability of linear systems. Numerical results showing the amount of reduction of fused track covariance by taking into account the effects of cross-correlation between candidate tracks for fusion is also presented     

高超声速飞行器单因素融合鲁棒预测控制

针对离线预测控制忽略各个因素对控制律的影响程度,将系统状态变量一维子空间对应的单因素控制量序列进行融合计算,设计了高超声速飞行器的单因素融合鲁棒预测控制器。针对系统的一维子空间离线生成单因素椭圆不变集序列和相应的控制量序列。在线时,在单因素椭圆不变集中分别搜索包含当前状态的最小椭圆,根据当前状态对控制序列进行融合得到控制律。仿真结果表明,相比于传统离线鲁棒预测控制算法,该方法能实现对高度和速度指令的快速、精确跟踪;相比于在线鲁棒预测控制算法,在线计算时间大大减少,实现了对高超声速飞行器的实时控制。     

Architectures and algorithms for track association and fusion

Target tracking using multiple sensors can provide better performance than using a single sensor. One approach to multiple target tracking with multiple sensors is to first perform single sensor tracking and then fuse the tracks from the different sensors. Two processing architectures for track fusion are presented: sensor to sensor track fusion, and sensor to system track fusion. Technical issues related to the statistical correlation between track estimation errors are discussed. Approaches for associating the tracks and combining the track state estimates of associated tracks that account for this correlation are described and compared by both theoretical analysis and Monte Carlo simulations     

基于伴随方程方法的非结构网格自适应技术及应用

对基于伴随方程的网格自适应技术的基本原理进行了研究,构造并实现了基于流场伴随方程的非结构网格自适应探测器,利用局部加密网格的办法,建立了基于Euler方程的无粘伴随自适应能力。对于粘性流动问题,提出首先应用无粘计算和伴随自适应技术从基准网格自动生成具有较好网格分辨率的非结构自适应网格,然后采用层推进法自动生成粘性计算网格的求解策略。采用RAE2822二维翼型和ONERA-M6三维机翼对所建立的无粘伴随自适应方法进行了验证,并应用所提出的方法和策略,对第二届国际涡流试验项目(VFE-2)中的65°尖前缘三角翼大攻角涡流场进行了数值模拟。对比计算结果表明,所建立的非结构网格伴随自适应方法是可行的和有效的。     

Joint sensor registration and track-to-track fusion for distributed trackers

Sensor registration deals with the correction of registration errors and is an inherent problem in all multisensor tracking systems. Traditionally, it is viewed as a least squares or a maximum likelihood problem independent of the fusion problem. We formulate it as a Bayesian estimation problem where sensor registration and track-to-track fusion are treated as joint problems and provide solutions in cases 1) when sensor outputs (i.e., raw data) are available, and 2) when tracker outputs (i.e., tracks) are available. The solution to the latter problem is of particular significance in practical systems as band limited communication links render the transmission of raw data impractical and most of the practical fusion systems have to depend on tracker outputs rather than sensor outputs for fusion. We then show that, under linear Gaussian assumptions, the Bayesian approach leads to a registration solution based on equivalent measurements generated by geographically separated radar trackers. In addition, we show that equivalent measurements are a very effective way of handling sensor registration problem in clutter. Simulation results show that the proposed algorithm adequately estimates the biases, and the resulting central-level trucks are free of registration errors.     

Sequence comparison techniques for multisensor data fusion andtarget recognition

A new class of techniques for multisensor fusion and target recognition is proposed using sequence comparison by dynamic programming and multiple model estimation. The objective is to fuse information on the kinematic state and “nonkinematic” signature of unclassified targets, assessing the joint likelihood of all observed events for recognition. Relationships are shown to previous efforts in pattern recognition and state estimation. This research applies “classical” speech processing-related and other sequence comparison methods to moving target recognition, extends the efforts of previous researchers through improved fusion with kinematic information, relates the proposed techniques to Bayesian theory, and applies parameter identification methods to target recognition for improved understanding of the subject in general. The proposed techniques are evaluated and compared with existing approaches using the method of generalized ambiguity functions, which lends to a form of Cramer-Rao lower bound for target recognition     

Comparison of two measurement fusion methods forKalman-filter-based multisensor data fusion

Currently there exist two commonly used measurement fusion methods for Kalman-filter-based multisensor data fusion. The first (Method I) simply merges the multisensor data through the observation vector of the Kalman filter, whereas the second (Method II) combines the multisensor data based on a minimum-mean-square-error criterion. This paper, based on an analysis of the fused state estimate covariances of the two measurement fusion methods, shows that the two measurement fusion methods are functionally equivalent if the sensors used for data fusion, with different and independent noise characteristics, have identical measurement matrices. Also presented are simulation results on state estimation using the two measurement fusion methods, followed by the analysis of the computational advantages of each method     

On optimal track-to-track fusion

Track-to-track fusion is an important part in multisensor fusion. Much research has been done in this area. Chong et al. (1979, 1986, 1990) among others, presented an optimal fusion formula under an arbitrary communication pattern. This formula is optimal when the underlying systems are deterministic, i.e., the process noise is zero, or when full-rate communication (two sensors exchange information each time they receive new measurements) is employed. However, in practice, the process noise is not negligible due to target maneuvering and sensors typically communicate infrequently to save communication bandwidth. In such situations, the measurements from two sensors are not conditionally (given the previous target state) independent due to the common process noise from the underlying system, and the fusion formula becomes an approximate one. This dependence phenomena was also observed by Bar-Shalom (1981) where a formula was derived to compute the cross-covariance of two track estimates obtained by different sensors. Based on this results a fusion formula was subsequently derived (1986) to combine the local estimates which took into account the dependency between the two estimates. Unfortunately, the Bayesian derivation made an assumption that is not met. This work points out the implicit approximation made and shows that the result turns out to be optimal only in the ML (maximum likelihood) sense. A performance evaluation technique is then proposed to study the performance of various track-to-track fusion techniques. The results provide performance bounds of different techniques under various operating conditions which can be used in designing a fusion system.     

UWB radar for human being detection

UWB radar for detection and positioning of human beings in complex environment has been developed and manufactured. Novelty of the radar lies in its large operational bandwidth (11.7GHz at -10dB level) combined with high time stability. Detection of respiratory movement of a person in laboratory conditions has been demonstrated. Based on experimental results human being radar return has been analysed in the frequency band from 1 GHz to 12 GHz. Novel principle of human being detection is considered and verified experimentally.     

Motion estimation for moving target detection

This paper describes a suite of techniques for the autonomous detection of moving targets by processing electro-optical sensor imagery (such as visible or infrared imagery). Specific application scenarios that require moving target detection capability are described, and solutions are developed under the constraints imposed by the scenarios. Performance evaluation results are presented using a test data set of over 300 images, consisting of real imagery (visible and infrared) representative of the application scenarios     

Hough transform for long chirp detection

The online detection of a very long and weak chirp signal is studied. The signal has an extremely slowly decreasing frequency, and is corrupted by white Gaussian noise and possibly also by powerful tones. By exploring and comparing candidate methods, it is found that the Hough transform (HT) detector appears to be most suitable given constraints on computational load and detectability. The analytical and the simulational performance of the HT detector are obtained and compared with the analytical performance of the generalized likelihood ratio test (GLRT), which is assumed to be optimal. Applying a suitable threshold for the HT can increase speed dramatically while preserving performance. We have found that both dithering (taking varied frequency shifts for fast Fourier transforms (FFTs)) and increasing the FFT length can reduce the minimum detectable frequency slope with nearly no additional computation