Alireza Moghaddamjoo

Road Extraction From Satellite Images Using Particle Filtering and Extended Kalman Filtering

Extended Kalman filter (EKF) has previously been employed to extract road maps in satellite images. This filter traces a single road until a stopping criterion is satisfied. In our new approach, we have combined EKF with a special particle filter (PF) in order to regain the trace of the road beyond obstacles, as well as to find and follow different road branches after reaching to a road junction. In this approach, first, EKF traces a road until a stopping criterion is met. Then, instead of terminating the process, the results are passed to the PF algorithm which tries to find the continuation of the road after a possible obstacle or to identify all possible road branches that might exist on the other side of a road junction. For further improvement, we have modified the procedure for obtaining the measurements by decoupling this process from the current state prediction of the filter. Removing the dependence of the measurement data to the predicted state reduces the potential for instability of the road-tracing algorithm. Furthermore, we have constructed a method for dynamic clustering of the road profiles in order to maintain tracking when the road profile undergoes some variations due to changes in the road width and intensity.

Robust adaptive Kalman filtering with unknown inputs

The conventional sequential adaptive procedure for estimating noise covariances and input forcing function has suboptimal performance and potential instability. In this work we present a robust procedure for optimally estimating a polynomial-form input forcing function, its time of occurrence and the measurement error covariance matrix, R. This procedure is based on a running window robust regression analysis. In addition a general robust procedure for estimating the process noise covariance matrix, Q, is derived. This procedure is based on the optimal filters residual characteristics and stochastic approximation.

Speckle Suppression in SAR Images Using the 2-D GARCH Model

A novel Bayesian-based speckle suppression method for Synthetic Aperture Radar ( SAR) images is presented that preserves the structural features and textural information of the scene. First, the logarithmic transform of the original image is analyzed into the multiscale wavelet domain. We show that the wavelet coefficients of SAR images have significantly non-Gaussian statistics that are best described by the 2-D GARCH model. By using the 2-D GARCH model on the wavelet coefficients, we are capable of taking into account important characteristics of wavelet coefficients, such as heavy tailed marginal distribution and the dependencies between the coefficients. Furthermore, we use a maximum a posteriori (MAP) estimator for estimating the clean image wavelet coefficients. Finally, we compare our proposed method with various speckle suppression methods applied on synthetic and actual SAR images and we verify the performance improvement in utilizing the new strategy.

Robust adapative Kalman filtering for systems with unknown step inputs and non-Gaussian measurement errors

Target tracking with Kalman filters is hampered by target maneuvering and unknown process and measurement noises. We show that moving data windows may be used to analyze state and measurement error sequences, determining robust estimates of bias and covariance. For steps in the system forcing functions and non-Gaussian measurement errors, the robust estimators yield improvements over linear bias and covariance estimators. Extensive simulations compare conventional, linear adaptive, and robust adaptive average step responses of a first-order system filter. Quantities examined are state estimate, state error, process and measurement covariance estimates, Kalman gain, and input step estimate.

Transform-based covariance differencing approach to the array with spatially nonstationary noise

The problem of bearing estimation in a situation where the additive white noise is assumed to be spatially nonstationary, i.e uncorrelated and of unequal power from sensor to sensor, is discussed. In this condition, the resolution of the standard methods decreases and the problem of source enumeration becomes more involved. A transform-based covariance differencing approach is proposed to resolve the problems of enumerating sources and estimating their bearings. Performance of the proposed method is compared with that of MUSIC via simulation. >

Two-dimensional DFT projection for wideband direction-of-arrival estimation

A statistical analysis of the spatial-temporal DFT projection technique for estimation of the direction-of-arrivals (DOAs) of wideband signals is presented. One of the advantages of this technique is its simple implementation as it does not require the construction of focusing matrices. Simulation results and comparisons with existing techniques show satisfactory performance, especially at low signal-to-noise (SNR) ratios and a relatively low SNR threshold. >

Application of spatial filters to DOA estimation of coherent sources

Problems associated with spatial smoothing which relate to the rate of decorrelation of correlated sources are discussed. Although a small improvement is attained, if sources have small differences in their directions-of-arrival (DOAs), they will remain highly correlated and the problem is not significantly diminished. A method which is based on the spatial filtering of an array vector is proposed. A procedure is developed that significantly reduces the correlations among sources and results in a nearly diagonal source covariance matrix. The design of the proposed spatial filters requires preliminary estimates of DOAs. This requirement, however, is not a drawback to this method, as is shown. Simulation studies are carried out for a 12-element linear array with quarter-wavelength element spacing. In all cases, 100 snapshots are used to estimate the array covariance matrix. In each case, 10 different simulations, with different noise sequences, are performed to present the variability of the resultant spectrum. The number of sources is set to 3 with -60 degrees , 30 degrees , and 40 degrees DOAs. >

Automatic segmentation and classification of ionic-channel signals

An automatic channel detection algorithm is proposed. This algorithm is based on sequential minimization of an index which is usually used in cluster analysis. The algorithm consists of two stages, namely segmentation and classification. In the first stage, the signal samples are segmented based on the assumption that the samples in each segment should be sequentially connected. In the second stage, the resultant segments are classified with no regard to their connectivities. The algorithm is computationally fast and globally optimum. The criterion function used in this algorithm is the ratio of the within-class variation over the between-class variation. An information-theoretic criterion that can be used mainly as a stopping rule in the segmentation stage is also proposed. Results on synthetic and real channel currents suggest that this algorithm will substantially increase the productivity of many laboratories involved in ionic-channel research.<<ETX>>

Spatial-Temporal DFT projection for wideband array processing

A new approach for high-resolution direction-of-arrival (DOA) estimation of coherent wideband signals is proposed. The approach is based on the projection of the two-dimensional spatial-temporal DFT of the received signal. The projections are used to construct narrowband covariance matrices having the same signal subspace which are then combined to yield the array covariance matrix. The array covariance can then be used by any signal-subspace eigendecomposition algorithm. The proposed approach does not require any preliminary estimates of DOAs and is not restricted to particular source models.<<ETX>>

A robust running-window detector and estimator for step-signals in contaminated Gaussian noise

An N-point window is applied to noisy data to recover stepped signals in non-Gaussian noise. Robust measures of signal step level and noise distribution spread are used to detect sequential clusters of data points which are statistically significantly different, thereby detecting the step. Using conventional analysis-of-variance methods, but with robust parameter estimates, false alarm probabilities are set reasonably accurately, and miss probabilities and signal level estimates are shown by simulation to yield good results. Applications to Kalman filtering, seismic and well-log data, and image processing are indicated.