Abstract: System control includes sensor management, user refinement, and mission accomplishment (SUM). An example of simultaneous tracking and identification includes (1) mission goals of resource appropriation and goal priorities, (2) user selection of targets and areas of coverage, and (3) fusion of data and sensory information. Many sensor management (SM) approaches are data-driven which includes filtering, aggregation, and normalization; however that does not include intelligent design. A top-down approach would facilitate the use of the right sensor, collecting the needed information, at the correct time. In order to better design SM algorithms, we utilize sensor, target, environmental, and automatic target recognition performance models for automatic target exploitation (ATE) prediction. Similar to pruning nodes in a Bayes net aggregation, a sensor manager can utilize the operating conditions (OCs) {i.e. sensor, target, environment} to condition the cost function, sensor-to-target assignment constraints, and scheduling times. An example is presented of determining task value of electro-optical sensor selection and scheduling based on the range to target, target size, and environmental conditions (e.g. occlusions). The key aspect of the SMOC provides accurate assignment and scheduling based on up-to-date database information, a capabilities matrix, and pragmatic sensor use to improve task satisfaction.

Abstract: This paper deals with the problem of automatic target recognition (ATR) of non cooperative targets. Specifically the focus is on the ATR of ships in ISAR images. To fulfill this task a multifeature based technique is proposed which uses a number of features extracted from the ship radar image. Both cases of known/unknown target aspect angle and single/multiframe based processing technique are considered in order to assess the performance improvement arising from the a priori knowledge of the aspect angle and from the joint use of the features extracted from a number of different frames. The performance of the proposed technique is investigated in depth by means of simulated data. Moreover results of its application to live ISAR images of ship targets are also provided showing the effectiveness of the proposed approach.

Abstract: A novel thresholding algorithm is presented in this paper to improve image segmentation performance at a low computational cost. The proposed algorithm uses a normalized graph-cut measure as thresholding principle to distinguish an object from the background. The weight matrices used in evaluating the graph cuts are based on the gray levels of the image, rather than the commonly used image pixels. For most images, the number of gray levels is much smaller than the number of pixels. Therefore, the proposed algorithm requires much smaller storage space and lower computational complexity than other image segmentation algorithms based on graph cuts. This fact makes the proposed algorithm attractive in various real-time vision applications such as automatic target recognition. Several examples are presented, assessing the superior performance of the proposed thresholding algorithm compared with the existing ones. Numerical results also show that the normalized-cut measure is a better thresholding principle compared with other graph-cut measures, such as average-cut and average-association ones.

Abstract: The concept of a novel forward scattering micro-radar wireless network for ground targets detection and recognition is presented The system topology and structure are described first, followed by a summary of the systempsilas capabilities and applications Signal processing strategies used for target detection, parameter estimation and automatic target recognition are briefly explained and supported with experimental results

Abstract: Parametric scattering center models match to radar scene attributes, aiding in automatic target recognition (ATR) and scene visualization. In this paper, we develop parametric models of canonical shapes for bistatic synthetic aperture radar (SAR).We generalize geometric theory of diffraction solutions for scattering mechanisms in a plane to develop three-dimensional models for six canonical shapes: a rectangular plate, dihedral, trihedral, cylinder, top-hat, and sphere. The proposed models provide physically relevant yet compact scattering solutions that are easily implemented for radar signal processing and ATR applications. The derived models are shown to agree with results obtained from high-frequency, asymptotic scattering simulations.

Abstract: In this paper, the radar target recognition is given by projected features of frequency-diversity RCS (radar cross section). The frequency diversity means signals are collected by sweeping the frequency of the incident illumination. Initially, the frequency- diversity RCS data from targets are collected and projected onto the PCA (principal components analysis) space. The elementary recognition of targets is efficiently performed on the PCA space. To achieve well separate recognition of targets, the features of the PCA space are further projected onto the LDA (linear discriminant algorithm) space. Simulation results show that accurate results of radar target recognition can be obtained by the proposed frequency- diversity scheme. In addition, the proposed frequency-diversity scheme has good ability to tolerate noise effects in radar target recognition.

Abstract: One important issue for Automatic Target Recognition (ATR) systems is to learn how robust the performance is under different scenarios. The quality of the input image sequence is a major factor affecting the ATR algorithm's ability to detect and recognize an object. If one can correlate the algorithm performance with different image quality measures, the recognition confidence can be predicted before applying ATR by predetermining the input image quality. In this paper, we address the utility of image quality measures and their correlations with performance failures of a principle component analysis (PCA) based ATR algorithm. Various image fusion approaches are examined to illustrate their abilities to improve ATR performance. Results show that the Shift Invariant Discrete Wavelet Transform (SiDWT) and Laplacian pyramid fusion schemes outperform other methods for improving the detection rate with the considered SAR images. Regression analysis is conducted to show that linear combinations of the selected image quality measures could explain about 60% of the variability in the non-detections of the ATR algorithm.

Abstract: Automatic target recognition (ATR) system performance over various operating conditions is of great interest in military applications. The performance of ATR system depends on many factors, such as the characteristics of input data, feature extraction methods, and classification algorithms. Generally speaking, ATR performance evaluation can be performed either theoretically or empirically. The theoretical evaluation method requires reasonably accurate underlying models for characterizing target/clutter data, which in many cases is unavailable. The empirical (experimental) evaluation method, on the other hand, needs a fairly large data set in order to conduct meaningful experimental tests. In this paper, we present experimental performance evaluation of ATR algorithms using the Moving and Stationary Target Acquisition and Recognition (MSTAR) data set. We conduct a comprehensive analysis of the ATR performance under different operating conditions. In the experimental tests, different feature extraction techniques, Principle Component Analysis (PCA), Linear Discriminant Analysis (LDA) and kernel PCA, are employed on target SAR imagery to reduce the feature dimension. A number of classification approaches, Nearest Neighbor, Naive Bayes, Support Vector Machine are tested and compared for their classification accuracy under different conditions such as various feature dimensions, target classes, feature selection methods and input data quality. Our experimental results provide a guideline for selecting features and classifiers in ATR system using synthetic aperture radar (SAR) imagery.

Abstract: Automatic target recognition (ATR) is an important capability for defense application. ATR removes the human operator from the process of target acquisition and classification, reducing the reaction time to possible threats and can be used to gun target engagement. This paper presents one technique used to solve the automatic target recognition problem in synthetic aperture radars (SAR) images, that is independent of target pose in the images. The classification is performed by a combination of three different classifiers the minimum distance classifier (MDC), the quadratic Gaussian classifier (QGC) and a multilayer perceptron (MLP) neural network, using a voting architecture.

Abstract: A microphone-array-based speech recognition system using a blind source separation (BBS) and a target speech extraction method in the system are provided. The speech recognition system performs an independent component analysis (ICA) to separate mixed signals input through a plurality of microphone into sound-source signals, extracts one target speech spoken for speech recognition from the separated sound-source signals by using a Gaussian mixture model (GMM) or a hidden Markov Model (HMM), and automatically recognizes a desired speech from the extracted target speech. Accordingly, it is possible to obtain a high speech recognition rate even in a noise environment.

Abstract: Due to the difficulty of creating training databases using all real enemy targets, it is necessary to derive them using computer simulations. In this paper, we apply three high frequency radar cross section (RCS) methods to create a training database for automatic target recognition (ATR) using 1-D range profiles. These methods are: physical optics (PO), physical theory of diffraction (PTD) and shooting and bouncing ray (SBR). Experimental results derived from the performance of combinational feature space trajectory with a new distance metric (FSTND) classifier show that PO+PTD is the most efficient method for ATR because of the additional information by diffraction terms. SBR shows poor performance due to the cavity structure.

Abstract: In this paper, we consider an urban ISR scenario where a human operator is tasked to provide feedback regarding the nature of some objects of interest. The feedback is relayed to the stochastic controller of an unmanned aerial vehicle (UAV), which must determine an appropriate mission plan. A small (unmanned) aerial vehicle (SAV) loiters at a high altitude where it may survey a large territory. An operator decides which objects in the SAV’s field of view are of interest and which are not. Then a team of micro (unmanned) aerial vehicles (MAVs) is assigned individual tours to inspect the objects of interest at a low altitude. As a MAV flies over an object of interest, the operator must decide if the object has a feature that uniquely distinguishes it as a target. The key parameters are the operator’s response and the time taken for the operator to respond. The controller uses these parameters to compute the expected information gain of a revisit. In previous studies automatic target recognition (ATR) was used for making some decisions in the SAV and the MAVs. This paper investigates the use of human feedback alone for target recognition. Different methods for calculating expected information gain are examined and compared. In addition, results from a flight test of this controller are presented.

Abstract: Automatic target recognition (ATR) is generally the reason why ISAR imaging systems are employed. Moreover, the use of fully polarimetric radar systems in radar imaging applications such as SAR and ISAR has enhanced both image quality and classification capabilities. In this paper, the authors propose a novel technique for ATR by using Polarimetric ISAR (Pol-ISAR) images. The proposed method is based on a model matching approach. Results are obtained by using real data that show the effectiveness of such technique.

Abstract: Laser-based 3D sensors measure range with high accuracy and allow for detection of several reflecting surfaces for each emitted laser pulse. This makes them particularly suitable for sensing objects behind various types of occlusion, e.g. camouflage nets and tree canopies. Nevertheless, automatic detection and recognition of targets in forested areas is a challenging research problem, especially since foreground objects often cause targets to appear as fragmented. In this paper we propose a sequential approach for detection and recognition of man-made objects in natural forest environments using data from laser-based 3D sensors. First, ground samples and samples too far above the ground (that cannot possibly originate from a target) are identified and removed from further processing. This step typically results in a dramatic data reduction. Possible target samples are then detected using a local flatness criterion, based on the assumption that targets are among the most structured objects in the remaining data. The set of samples is reduced further through shadow analysis, where any possible target locations are found by identifying regions that are occluded by foreground objects. Since we anticipate that targets appear as fragmented, the remaining samples are grouped into a set of larger segments, based on general target characteristics such as maximal dimensions and generic shape. Finally, the segments, each of which corresponds to a target hypothesis, undergo automatic target recognition in order to find the best match from a model library. The approach is evaluated in terms of ROC on real data from scenes in forested areas.

Abstract: Multi-source data fusion in the context of automatic target recognition (ATR) involves the fusion of multiple, independent observations of a phenomenon. If the collection of sources is diverse, the resulting classification system is expected to perform better than one based on any one source. In recent work, the authors have demonstrated the use of such decision fusion strategies in alleviating the over-dimensionality and small-sample-size problems associated with hyperspectral data. Multi-temporal hyperspectral recognition and classification tasks are even more prone to over-dimensionality of features and small training sample size problems. In this work, the authors will extend their previously proposed framework to multi-temporal, hyperspectral target recognition / classification problems. The performance of the proposed system will be compared against that of conventional hyperspectral feature extraction techniques. The efficacy of the proposed system is quantified by overall recognition accuracies.

Abstract: This paper describes the automatic target recognition (ATR) challenge problem which includes source code for a baseline ATR algorithm, display utilities for the results, and a high range resolution (HRR) data set consisting of 10 civilian vehicles. The Ku-band data in this data set has been processed into 1-dimensional range profiles of vehicles in the open, moving in a straight line. It is being released to the ATR community to facilitate the development of new and improved HRR identification algorithms which can provide greater confidence and very high identification performance. The intent of the baseline algorithm included with this challenge problem is to provide an ATR performance comparison to newly developed algorithms. Single-look identification performance results using the baseline algorithm and the data set are provided as a starting point for algorithm developers. Both the algorithm and data set can support single look and multi-look target identification.

Abstract: Inverse Synthetic Aperture Radar (ISAR) images are often used for classifying and recognising targets. Moreover the use of a fully polarimentric ISAR image enhances classiication capabilities. In this paper, the authors propose a novel ATR technique based on the use of fully polarimetric ISAR images and Neural Networks. In order to reduce the amount of data processed by the classifier, the brightest scattering centres are first extracted by means of the Pol-CLEAN technique and then their scattering matrices are decomposed using Cameron's decomposition. The proposed ATR algorithm is finally tested on real data.

Abstract: This paper presents a novel fusion technique for automatic target recognition from high range resolution radar profiles when observations from multiple viewpoints are available. The fusion technique entails only a straightforward modification of the transition probabilities of a single-viewpoint target model in which a Hidden Markov Model is used to represent the unknown target orientation. Evaluations using the MSTAR database indicate that the new technique can reduce classification errors by about two orders of magnitude when compared to single viewpoint observations and, in a 10-target classification experiment, gave almost perfect recognition.

Abstract: An automatic target recognition system with adaptive metric selection. The novel system includes an adaptive metric selector for selecting a match metric based on the presence or absence of a particular feature in an image and a matcher for identifying a target in the image using the selected match metric. In an illustrative embodiment, the adaptive metric selector is designed to detect a shadow in the image and select a first metric if a shadow is detected and not cut off, and select a second metric otherwise. The system may also include an automatic target cuer for detecting targets in a full-scene image and outputting one or more target chips, each chip containing one target. The adaptive metric selector adaptively selects the match metric for each chip separately, and may also adaptively select an appropriate chip size such that a shadow in the chip is not unnecessarily cut off.

Abstract: Automatic target recognition (ATR) is generally the reason why ISAR imaging systems are employed. Moreover, the use of fully polarimetric radar systems in radar imaging applications such as SAR and ISAR has enhanced both image quality and classification capabilities. In this paper, the authors propose a novel technique for ATR by using Polarimetric ISAR (Pol-ISAR) images. The proposed method is based on a model matching approach. Results are obtained by using real data that show the effectiveness of such technique.

Abstract: In this paper the manifold geometry in radar high-resolution range profile (HRRP) is firstly explored, and then, according to the characteristics of target pose sensitivity, a method of adaptively segmenting the aspect sectors is proposed for HRRP recognition through evaluating the curvature of HRRP manifold. Promising experimental results are presented for measured radar data.

Abstract: This paper primarily investigates the use of shape-based features by an Automatic Target Recognition (ATR) system to classify various types of targets in Synthetic Aperture Radar (SAR) images. In specific, shapes of target outlines are represented via Elliptical Fourier Descriptors (EFDs), which, in turn, are utilized as recognition features. According to the proposed ATR approach, a segmentation stage first isolates the target region from shadow and ground clutter via a sequence of fast thresholding and morphological operations. Next, a number of EFDs are computed that can sufficiently describe the salient characteristics of the target outline. Finally, a classification stage based on an ensemble of Support Vector Machines identifies the target with the appropriate class label. In order to experimentally illustrate the merit of the proposed approach, SAR intensity images from the well-known Moving and Stationary Target Acquisition and Recognition (MSTAR) dataset were used as 10-class and 3-class recognition problems. Furthermore, comparisons were drawn in terms of classification performance and computational complexity to other successful methods discussed in the literature, such as template matching methods. The obtained results portray that only a very limited amount of EFDs are required to achieve recognition rates that are competitive to well-established approaches.

Abstract: We propose a novel target recognition algorithm for classification of three types of ground vehicles in the moving and stationary target acquisition and recognition public release database. Algorithms that produce classifiers with large margins, such as support vector machines (SVMs), AdaBoost, etc. are receiving more and more attention in the literature. A real application of AdaBoost for synthetic aperture radar automatic target recognition is presented and the result is compared with conventional classifiers. And we also describe how AdaBoost algorithm can be used as a multiclass classification method as well as a feature fusion method. Results are presented to verify that, the performance of the recognition system is improved significantly, and the method presented in this paper is an effective method for SAR images feature fusion and target recognition.

Abstract: An improved automatic target recognition (ATR) processing string has been developed. The overall processing string consists of pre-processing, subimage adaptive clutter filtering (SACF), normalization, detection, data regularization, feature extraction, optimal subset feature selection, feature orthogonalization and classification processing blocks. A new improvement was made to the processing string, data regularization, which entails computing the input data mean, clipping the data to a multiple of its mean and scaling it, prior to feature extraction. The classified objects of 3 distinct strings are fused using the classification confidence values and their expansions as features, and using "summing" or log-likelihood-ratio-test (LLRT) based fusion rules. The utility of the overall processing strings and their fusion was demonstrated with new high-resolution three-frequency band sonar imagery. The ATR processing strings were individually tuned to the corresponding three-frequency band data, making use of the new processing improvement, data regularization, which resulted in a 3:1 reduction in false alarms. Two significant fusion algorithm improvements were made. First, a nonlinear 2nd order (Volterra) feature LLRT fusion algorithm was developed. Second, a repeated application of a subset Volterra feature selection / feature orthogonalization / LLRT fusion block was utilized. It was shown that cascaded Volterra feature LLRT fusion of the ATR processing strings outperforms baseline summing and single-stage Volterra feature LLRT algorithms, yielding significant improvements over the best single ATR processing string results, and providing the capability to correctly call the majority of targets while maintaining a very low false alarm rate.

Abstract: A recognizing and positioning method of ground buildings belongs to the imaging automatic target recognition field, which aims at solving the problem of recognition and positioning from different viewpoints and from different scales, and different heights, and to be used in forward looking ground buildings. The invention constructs a ground building standard feature library in advance. The sequence includes: an enhanced image procedure, a background suppression processing procedure, a gray level merge procedure, a feedback and segmentation procedure, a vertical bar feature detection procedure, and a quadratic character matching procedure. The invention further extracts characteristic quantity to match with the standard characteristic, considers recognizing the veins and scene information of the buildings, and recognizes and positions the forward looking ground buildings, aiming at the characteristic of ground buildings, and making use of mathematical morphology to extract structure information of image. The novel method has high precision of recognition, good reliability, and is used in fields such as urban planning, supervision, aircraft contact navigation, collision-avoidance to recognize the forward looking ground buildings of different viewpoints, different scales and different heights.