Recursive 3-D motion estimation from a monocular image sequence

Consideration is given to the design and application of a recursive algorithm to a sequence of images of a moving object to estimate both its structure and kinematics. The object is assumed to be rigid, and its motion is assumed to be smooth in the sense that it can be modeled by retaining an arbitrary number of terms in the appropriate Taylor series expansions. Translational motion involves a standard rectilinear model, while rotational motion is described with quaternions. Neglected terms of the Taylor series are modeled as process noise. A state-space model is constructed, incorporating both kinematic and structural states, and recursive techniques are used to estimate the state vector as a function of time. A set of object match points is assumed to be available. The problem is formulated as a parameter estimation and tracking problem which can use an arbitrarily large number of images in a sequence. The recursive estimation is done using an iterated extended Kalman filter (IEKF), initialized with the output of a batch algorithm run on the first few frames. Approximate Cramer-Rao lower bounds on the error covariance of the batch estimate are used as the initial state estimate error covariance of the IEKF. The performance of the recursive estimator is illustrated using both real and synthetic image sequences     

Reciprocal radar waveforms

Digitally coded radar waveforms can be used to obtain large time-bandwidth products (pulse compression ratios). It is demonstrated that periodic radar waveforms with zero sidelobes or almost zero sidelobes can be defined. A perfect periodic code is a periodic code whose autocorrelation function has zero sidelobes and whose amplitude is uniform (maximum power efficiency=1). An asymptotically perfect periodic code has the property that as the number of elements in the code goes to infinity the autocorrelation function of the code has zero sidelobes and its power efficiency is one. The authors introduce a class of radar waveforms that are either perfect or asymptotically perfect codes. These are called reciprocal codes because they can be derived through a linear transformation of known codes. The aperiodic performance of the reciprocal code is examined     

Chirp Waveform Generation Using Digital Samples

This paper describes two methods of generating an analog frequency-modulated waveform by the use of a small number of digital samples of the ?chirp? waveform. The number of digital samples required is a function of the time-bandwidth product. For certain values of time-bandwidth product, this type of signal generation becomes extremely efficient. Several proofs are offered which show how to select ?optimum? values of time-bandwidth products. Two hardware implementations are suggested. One is based on the use of modulo arithmetic and a small stored memory table. The second method utilizes the inherent signal symmetries available if ?optimum? time-bandwidth products are selected. The symmetrical signal patterns are stored in recirculating reversible shift registers which can be read out at high speeds.     

Multitarget detection using Kalman-based average displacementestimator

The anomaly in the displacement estimates obtained from a low-level Kalman-based average displacement estimator is used to detect multiple targets in a forward-looking infrared (FLIR) scene. The displacement estimates originating from the targets could be associated with the detected multiple targets and an updated estimated position of the target could be obtained. This procedure could be repeated for all targets in the scene to obtain multitarget tracking. The behavior of the expected value of the displacement estimates as a function of the number of iterations is investigated. The behavior of the displacement error covariance matrix and the Kalman gain matrix are discussed as functions of the number of iterations     

Phase-Coded Interrupted CW Signals and Processors

High resolution radars require signals with large time-bandwidth product such as CW signal and coherent pulse train (CPT). We discuss a phase-coded interrupted CW (ICW) signal which is the combination of CW signal and CPT. Phase codes used here are with perfect periodic autocorrelation. The periodic ambiguity function of ICW signals is studied including single-carrier signal and multi-carrier signal. It is interesting that the gate function has different effects on two signals and contributes to a multi-carrier ICW signal which yields nearly perfect autocorrelation. Meanwhile we also suggest an efficient receiver approach to ICW signals, which can reduce the computational burden of the processor and utilize the good properties of P3 and P4 codes.     

A Kalman Tracker with a Simple Input Estimator

Two Kalman filter based schemes are proposed for tracking maneuvering targets. Both schemes use least squares to estimate a target's acceleration input vector and to update the tracker by this estimate. The first scheme is simpler and by an approximation to its input estimator the computation can be considerably reduced with insignificant performance degradation. The second scheme requires two Kalman filters and hence is more complex. However, since one of its two filters assumes input noise, it may outperform the first scheme when input noise is indeed present. A detector that compares the weighted norm of the estimated input vector to a threshold is used in each scheme. Its function is to guard against false updating of the trackers and to keep the error covariance small during constant velocity tracks. Simulation results for various target profiles are included. They show that in terms of tracking performance, both schemes are comparable. However, because of its computation simplicity, the first scheme is far superior.     

Time-Frequency HOP Signals Part I: Coding Based upon the Theory of Linear Congruences

Time-frequency hop codes are developed based upon the theory of linear congruences. These codes can be used for multiuser radar and asynchronous spread spectrum communications systems. A uniform upper bound is placed on the cross-correlation function between any two elements of the code set. The upper bound is minimized by choice of time-bandwidth product and is shown to diminish as 2/N, where N is the number of elements in the code set. The size and position of spurious peaks in the autocorrelation functions are discussed. The results are extended to narrowband ambiguity functions.     

Maximum Likelihood Energy Detection of Mary Orthogonal Signals

The literature on energy detection is extended by applying it to the processing of M'ary orthogonal communication signals of arbitrary time-bandwidth product. Guassian noise channel transition probabilities are derived for maximum likelihood energy detection, modified to the extent of including an erasure threshold. Relations and computational techniques are described for determining the symbol erasure and error probabilities for general signal alphabet sizes (M) and time-bandwidth products. When time of arrival is known exactly and Doppler is negligible, gible, it is determined that energy detection is inferior to noncoherent matched filter detection. For time-bandwidth products of the order of 100 and error probabilities around 10 3, a loss of about 5 dB occurs which is attributable to a lack of knowledge of the detail signal structure. However, for problems where time of arrival and/ or Doppler are unknown, energy detection will perform nearly as well as matched filter detection of, for example, spread spectrum signals, and is also simpler to implement. General energy detection performance curves are given in terms of required signal energy for specific error and erasure probabilities, as a function of M'ary and time-bandwidth product.     

Linear FM Signal Formats for Beacon and Communication Systems

This paper examines the capabilities of the class of linear FM spread-spectrum signals within the context of potential communications systems usage in order to establish some performance criteria and bounds that permit comparison with other spread-spectrum formats. A systematic basis is provided for parameter selection for this class of signals by examining the interaction a mong the frequency-modulation indices, time-bandwidth product, and cross-talk criteria that determine the number of effective linear FM signals (or channels) that can be used within the constraints of a bounded time-frequency region. A general expression is derived relating N, the number of useful signals, R2, a cross-talk parameter, ToWo, the mean time-bandwidth product, and ?max and ?min, the maximum and minimum FM rates of the signal set. Canonic signal processor structures are described for ensembles of linear FM signals that have either constant duration or constant bandwidth. It is then shown that the signal modulation format can be modified in accordance with classical paired-echo theory to expand the utility of this class of signals in both synchronous and nonsynchronous operations to yield the equivalent of time-division and code multiplexing. Possible applications for this signal format are discussed.     

A Time Sidelobe Reduction Technique for Small Time-Bandwidth Chirp

A filtering technique is described which permits time sidelobe suppression of small time-bandwidth product shirp. Signal-to-noise loss and time sidelobe level sensitivity to Doppler shift are presented for a filtered chirp waveform whose time-bandwidth product is 10.     

Radiometric detection of spread-spectrum signals in noise ofuncertain power

The standard analysis of the radiometric detectability of a spread-spectrum signal assumes a background of stationary, white Gaussian noise whose power spectral density can be measured very accurately. This assumption yields a fairly high probability of interception, even for signals of short duration. By explicitly considering the effect of uncertain knowledge of the noise power density, it is demonstrated that detection of these signals by a wideband radiometer can be considerably more difficult in practice than is indicated by the standard result. Worst-case performance bounds are provided as a function of input signal-to-noise ratio (SNR), time-bandwidth (TW) product and peak-to-peak noise uncertainty. The results are illustrated graphically for a number of situations of interest. It is also shown that asymptotically, as the TW product becomes large, the SNR required for detection becomes a function of noise uncertainty only and is independent of the detection parameters and the observation interval     

Noise subspace techniques in non-gaussian noise using cumulants

We consider noise subspace methods for narrowband direction-of-arrival or harmonic retrieval in colored linear non-gaussian noise of unknown covariance and unknown distribution. The non-gaussian noise covariance is estimated via higher order cumulants and combined with correlation information to solve a generalized eigenvalue problem. The estimated eigenvectors are used in a variety of noise subspace methods such as multiple signal classification (MUSIC), MVDR and eigenvector. The noise covariance estimates are obtained in the presence of the harmonic signals, obviating the need for noise-only training records. The covariance estimates may be obtained nonparametrically via cumulant projections, or parametrically using autoregressive moving average (ARMA) models. An information theoretic criterion using higher order cumulants is presented which may be used to simultaneously estimate the ARMA model order and parameters. Third- and fourth-order cumulants are employed for asymmetric and symmetric probability density function (pdf) cases, respectively. Simulation results show considerable improvement over conventional methods with no prewhitening. The effects of prewhitening are particularly evident in the dominant eigenvalues, as revealed by singular value decomposition (SVD) analysis     

受地效影响飞机起飞着陆运动模型的参数辨识

利用基于最小模型误差法和线性不连续跳跃多重打靶法建立的非线性辨识法，辨识了飞机起飞着陆过程的非线性动态模型。对于包含复杂非线性项的动态系统，本方法可以从实际试验测量的系统非线性数据，确定飞机处于地面效应影响运动过程的系统模型，而不需要预先详细描述系统的非线性形式。算例表明该方法对于原始近似动态系统的状态估计是足够精确的。     

Application of Maximum Likelihood Estimation of Persymmetric Covariance Matrices to Adaptive Processing

The optimum weights for an adaptive processor are determined by solving a particular matrix equation. When, as is usually true in practice, the covariance matrix is unknown, a matrix estimator is required. Estimating the matrix can be computationally burden some. Methods of decreasing the computational burden by exploiting persymmetric symmetries are discussed. It is shown that the number of independent vector measurements required for the estimator can be decreased by up to a factor of two.     

MMAE-Based Control with Space-Time Point Process Observations

A multiple model adaptive estimator (MMAE) has been formulated to estimate the state of a dynamic system modeled by a linear stochastic differential equation, from which measurements, described as a noise-corrupted space-time point process functionally related to that state, are extracted. Assumed certainty equivalence is used to combine such an estimator with the LQ full-state feedback controller to synthesize a practical, implementable controller. Performance of the estimator and resultant controller characteristics are investigated via simulation as a function of approximation method used to limit the full-scale estimator to finite dimensionality and also as a function of important parameters defining the dynamics and observation processes.     

Stability of the Decentralized Estimation in the JTIDS Relative Navigation

The joint tactical information distribution system (JTIDS) includes a relative navigation capability with which each member can estimate its own position by utilizing the broadcast position estimates of other members plus the measured times of arrival of the radio signals. Previous simulation studies of the system have shown that unrestricted ranging between members can produce unstable navigation errors, but a covariance-based hierarchy may be stable. We have succeeded in proving, analytically, the stability of a closely related organization?a fixed-rank hierarchy. For all ownstate organizations we present time-domain equations for computing estimation-error covariance and for analyzing stability. A simple simulation is able to demonstrate the stability issues.     

Time-Frequency HOP Signals Part II: Coding Based upon Quadratic Congruences

High-efficiency multicomponent signals for maximization of signalto-noise ratio are investigated. Maximization of signal-to-noise ratio in colored noise requires control of volume distribution of the signal ambiguity function and transmission of unity efficiency signals. Signal efficiency is defined as the ratio of average power to the peak power. It is concluded that the signals must be frequency hop pulse trains. Quadratic congruences are chosen to place the components in time-frequency space. The number-theoretic properties of these signals provide bounds on the position and amplitude of the various peaks of the signal ambiguity function. The tradeoffs are shown between volume removal, number of component signals, and the time-bandwidth product.     

基于最近邻的协同定位模型研究

协同定位是多平台编队中的关键问题之一,是实现无人机、舰艇编队等定位控制的基础.从信息融合的角度研究了编队协同中的导航定位问题,提出了一种新的协同定位算法.推导了二维情况下,基于最近邻准则确定伪测量和相伴误差协方差矩阵的模型.仿真分析表明,该算法可以稳定地完成己平台运动要素的估计.     

Likelihood-ratio detection of frequency-hopped signals

The optimum system for intercepting frequency-hopped signals uses a channelized receiver and a likelihood-ratio test (LRT). Previous results on the performance of the optimum system have been based on Gaussian assumptions, which are generally valid for transmissions having large time-bandwidth areas. Results, obtained by Monte Carlo simulation, for relatively small time-bandwidth transmissions are given here. The signal model is a simple one known as “pure frequency hopping.” Comparisons with the LRT show that energy detection loss increases when the time-bandwidth product of the transmission is increased by increasing the number of frequencies, even when the number of pulses is also increased. The loss decreases when only the number of pulses is increased. Over the parameter range observed, binary detection loss tends to increase with the number of pulses and decrease with the number of frequencies. Results are included for a moving-window version of the LRT. A parameter of the LRT is the signal-to-noise ratio (SNR). The effect of using a design value not equal to the true SNR is shown     

无需辅助数据的分布式目标自适应检测器

在非高斯背景和没有辅助数据的条件下,研究了高分辨率雷达分布式目标的自适应检测问题.首先采用有序检测理论和协方差矩阵的迭代估计方法粗略估计散射点集合,进一步利用迭代估计方法获得协方差矩阵的近似最大似然估计,提出了无需辅助数据的自适应检测器(ADWSD).ADWSD在非高斯背景下具有近似恒虚警率特性,且检测性能远好于修正的...