This PDF file contains the editorial “Editorial: Ethical Practices” for OE Vol. 31 Issue 12

This PDF file contains the editorial “Guest Editorial: Automatic Target Recognition” for OE Vol. 31 Issue 12

The Self Adaptive Hierarchical Target Identification and Recognition Neural Network (SAHTIRN™) is a unique and powerful combination of state-of-the-art neural network models for automatic target recognition applications. It is a combination of three models: (1) an early vision segmentor based on the Canny edge detector, (2) a hierarchical feature extraction and pattern recognition system based on a modified Neocognitron architecture, and (3) a pattern classifier based on the backpropagation network. Hughes has extensively tested SAHTIRN™ with several ground vehicular targets using terrain board modeled IR imagery under a current neural network program sponsored by the Defense Advanced Research Projects Agency. In addition, extensive testing was conducted using several real IR and handwritten character databases. Hughes has demonstrated successful performance with 91 to 100% probability of correct classification over this wide variety of data. End-to-end system results from these experiments are provided and interim results from each stage of the SAHTIRN™ system are discussed.

The focus of the artificial neural vision learning (ANVIL) program is to apply neural network technologies to the air-to-surface 3-D automatic target recognition (ATR) problem. The 3-D multiple object detection and location system (MODALS) neural network was developed under the ANVIL program to simultaneously detect, locate, segment, and identify multiple targets. The performance results show a very high dentification accuracy, a high detection rate, and a low false alarm rate, even for areas with high clutter and shadowing. The results are shown as detection/false alarm curves and identification/false alarm curves. In addition, positional detection accuracy is shown for various scale sizes. To provide data for the program, visible terrain board imagery was collected under a variety of background and lighting conditions. Tests were made on more than 500 targets of five types and two classes. These targets varied in scale by up to − 25%, varied in azimuth by up to 120 deg, and varied in elevation by up to 10 deg. The performance results are shown for targets with resolution ranging from 9 to 700 pixels on target.

We describe an experimental model-based automatic target recognition (ATR) system, called XTRS, for recognizing 3-D vehicles in real or synthetic, ground-based or airborne, 2-D laser-radar range and intensity images. The key to recognition is a new generic matching engine that compares image events (e.g., silhouettes) and their constituent primitives to some appearance-model (AM) hierarchy, which describes how 3-D, possibly articulated, objects appear in the imagery. We describe each of the system's components, i.e., the event characterization (for extracting events from images and decomposing the events into primitives), the models, the matching, and the control. XTRS may also perform low-level information fusion by constructing feature-indicating interest images and combining them into an overall interest image used for locating an event. Examples of processing and performance results are given for real CO₂ ground-based laser-radar imagery. We also report on a performance evaluation of XTRS based on synthetic data where important parameters (e.g., range to target) are systematically varied. Overall, more than 1500 range and intensity image pairs have been used throughout the development of XTRS.

The echoes reflected by a sound ping emerging from an active sonar when it interacts with a target in its path can be remotely sensed by a receiver. The presented approach capitalizes on an inverse scattering method that exploits the presence of certain resonance features in the echoes returned by targets to classify them. Classifying underwater objects is important to naval programs such as mine countermeasures (MCM) and anti-submarine warfare (ASW) to preclude wasting of ordnance on false targets. Although the classification of complex shapes is still aformidable task, considerable progress has been made in classifying simple shapes such as spheroidal or cylindrical shells. This methodology has emphasized the extraction, isolation, and labeling of resonance features hidden within the echo, but little has been said yet about how these could be used to classify a target. A couple of simple examples illustrate exactly howthese resonances can be linked to the physical characteristics of the target, allowing for its unambiguous characterization. The procedure, although illustrated with active acoustics (i.e., sonar), can be extended to any active return from any sensor, including radar.

Acoustic signals scattered from submerged elastictargets have produced interesting backscattered signals in the appropriate frequency region particularly because of the presence of resonances. Considerable work has been done for spherically and cylindrically shaped objects out to moderate frequencies for both solids and shells. Only recently, however, have researchers reported results for elongated objects such as spheroidal solids and shells. Further, the higher-frequency region for spheres has not been investigated extensively. A series of calculations that are characteristic of elongated targets or of spherical shells in the higher-frequency region are reported. Some of the discussed results involve bending resonances caused by obliquely incident plane waves on elongated objects and their relation to flexural resonances of beams; classes of resonances that are related to waves creeping along the longest and shortest meridians of a spheroid; flexural resonances (A₀ resonances) for shells in the time domain near the coincidence frequency; and high-frequency thickness effects when scattering from shells, and their prediction based on flat plate theory. Several numerical examples are shown.

The multidimensional inverse scattering of objects buried in an inhomogeneous elastic background structure is studied. The entire medium, comprising inclusion and background, is probed by an ultrasonic force, and the field scattered by the object is observed along a receiver array. The goal is to retrieve both the geometry (imaging problem) and the constitutive parameters (inverse problem) of the object through an appropriate multiparameter direct linear inversion. Because the multidimensional inverse scattering problem is nonlinear and ill-posed, it is linearized within the Born approximation for inhomogeneous background, and a minimum-norm least-squares solution to the discretized version of the vector integral formulation is sought. The solution is based on a singular value decomposition of the forward operator matrix. A priori information can be incorporated into the algorithm to enhance the accuracy and improve the resolution of the recovered object characteristics. The method is illustrated on a 2-D problem where constrained least-squares inversion of the object is performed from synthetic data. A Tikhonov regularization scheme is also examined and compared to the minimum-norm least-squares estimate.

A new technique for recognizing two-dimensional objects independent of scale and orientation is presented. This technique's performance on real imagery of tactical military targets, both occluded and nonoccluded, is evaluated. The robustness of this method with respect to partial occlusion is shown. The relatively small storage requirements and fast search time make it an attractive candidate for real-time applications.

An ongoing study of automatic target recognition using a passive multisensor suite is presented. The multisensor suite consists of electro-optical and acoustic sensors, a radio frequency interferometer, and meteorological sensors. Three feature analysis tools-a genetic algorithm, a branch-and-bound algorithm, and a new class overlap region partitioning scheme (CORPS)-are employed for assessing the effectiveness of features extracted by several data and feature fusion concepts. Dedicated hardware has been configured for laboratory and field testing. A high-resolution graphics display has been designed for monitoring the target-acquisition-related processes and maximizing information transfer to the operator.

The problem of region segmentation is examined and a new algorithm for maximum a posteriori (MAP) segmentation is introduced. The observed image is modeled as a composite of two processes: a high-level process that describes the various regions in the images and a low-level process that describes each particular region. A Gibbs-Markov random field model is used to describe the high-level process and a simultaneous autoregressive random field model is used to describe the low-level process. The MAP segmentation algorithm is formulated from the two models and a recursive implementation forthe algorithm is presented. Results of the algorithm on various synthetic and natural textures clearly indicate the effectiveness of the approach to texture segmentation.

The applicability of discriminant-function-based feature selection techniques to the design of radar target classification systems is addressed. The characteristics and limitations of several nonparametric feature selection criteria based on discriminant functions are discussed from the point of view of developing systems for classifying targets based on their real-aperture radar returns. The results of applying feature selection criteria to a radar target database are presented. The classification performance of the selected feature sets is evaluated by means of Monte Carlo simulation studies.

The development of an information-theoretic image measure for sensor evaluation, under contract to the United States Air Force, is described. Although current approaches are based on human perception models, a need exists for evaluation of sensors for automatic target cuing/automatic target recognition (ATC/ATR) systems. Such an evaluation should be performed in terms of the probabilities of detection/identification and false alarms, independent of the idiosyncrasies of the specific ATC/ATR algorithms. Such an approach based on the information-theoretic content of images for the target versus background separability is being developed and applied to evaluating sensors using the tower test data collected at the Wright Laboratories.

Automatic object detection is one of the basic tasks performed by an image understanding system. Object detection approaches need to perform accurately and robustly over a wide range of scenes. Although a number of detection approaches have been developed and reported, a need remains for standards by which to judge the relative merits of such approaches. Image characteristics, object characteristics, and detection methodology are recognized as the main variables affecting object detection. A basis for their quantitative analysis and evaluation is developed. This research keeps object detection methodology constant while varying image and object characteristics to develop a set of quantiative standards. This requires an ability to derive a quantitative measure for the "clutter" observed in an image. A performance index for object detection approaches, as a function of scene nature, is valuable. Current approaches to image clutter or quality characterization are studied and a new measure based on image texture content and object characteristics is proposed. An extensive set of experimental studies is utilized to evaluate this texture-based image clutter (TIC) measure. TIC is shown to be better suited than other reported clutter measures because of its ability to accurately quantify perceptual effects and to serve as a robust indicator ofthe object detection and false alarm rates as a function of image clutter.

ZnSe axicon is used to efficiently generate approximations to diffraction-free Bessel function beams at λ = 10.6 μm. The central spot radius of ~ 170 μm generated when the axicon is illuminated by a plane wave is maintained for distances of up to 11 cm with sufficient laser power to allow drilling of plexiglass to be demonstrated.

A compact, parallel array of three Nd:Cr:GSGG laser rods is used to construct a quasi-cw laser. The array is pumped by concentrated solar light and is mounted in a single solar concentrator. The three laser rods use a common pair of laser mirrors to define the optical resonator. The three laser beams are not coherently coupled in these experiments. The simplicity of the design, and its reasonable stability in terms of vibration and optical misalignment, suggest that the design may be scalable for higher power.

Two new analysis techniques used to characterize a non-diffraction-limited pulsed laser system whose focal spot is too intense to record photographically and whose pointing varies in time are described. The data to be analyzed were collected by moving the focal point in three dimensions and recording the fraction of the incident energy that passed through a pinhole. Analyzing the data in two different ways provides two independent estimates of the properties of the laser beam that are in reasonable agreement. One method estimates the average size of the focal volume over the measurement period, whereas the other estimates its minimum size. The difference between the average and the minimum then indicates the range of variation in size. Two different estimates of the random motion of the beam are also obtained. Both analysis methods employ statistics to account for pulse-to-pulse variations in the size, location, and total energy of the pulses. Beam widths at different points along the beam axis can then be fit with the M2 model, which describes beams that are not diffraction limited, to provide a simple expression for the irradiance at any point in the focal volume.

A variety of technologies to improve optical performance in photolithography are surveyed. The technologies are classified according to which component of the optics they work on: the source, the mask, the pupil, or the image. Emphasis is put on, first, the advantages and disadvantages of annular illumination, which are studied by lithography simulation. Second, a polarized mask is proposed. The effect of polarization on imaging, which has not been explored yet, is studied and the result is that TE polarization produces images of the best quality. Third, as a part of pupil function, aberration optimization is discussed. The conclusion is that positive spherical aberration is suitable for projection optics.

Small angle measurement using a modified moiré interference technique is reported. The range of angle measurement for 25-μm pitch gratings is about ± 25 μrad and the accuracy of measurement is of the order of 0.2 μrad with a time constant of 0.1 s. A piezoelectric transducer (PZT) is used for angular displacement. The angular sensitivity under the present experimental conditions is found to be of the order of 21.8 mV/μrad.

Resolution-variable moiré topography for measuring the three-dimensional profile of an object is described. With this method, moiré fringes are formed by projecting two sets of interference fringes of laser beams on an object. The interference fringes are formed using a Mach-Zehnder interferometer and are divided into two sets by a beamsplitter. The image, including the moiré fringes, which are formed in accordance with the object depth, is detected by an image sensor. The effectiveness of this method is demonstrated by practically measuring the profiles of a small object. The intervals between adjacent moiré fringes could be experimentally changed from 0.16 to 1.6 mm. The advantage ofthis method is that the interval between moiré fringes can be easily changed continuously by a mechanical operation.

The principles of magneto-optical (MO) recording and readout are reviewed and issues of thin film media design for MO disks are discussed. Essential criteria for improved media performance, and in particular, concepts of magneto-optical interactions in relation to multilayer designs for enhancing the polarization rotation in erasable magnetooptical media are considered. Examples of media design for improved performance are given. The role of sharp reflection edges such as those occurring near plasma edge resonances in metals is discussed. It is shown that rare earth hexaboride (RB₆) reflector layers increase the polar Kerr rotation angle of MO media. Iso-Kerr rotation contours and iso-reflectance contours are developed to determine optimizing material properties for MO media design.

The probability of correct object recognition is calculated assuming a maximum-likelihood algorithm, accounting for clutter and for the opticaltransferfunction (OTF) of both the atmosphere and the imaging system. The OTF is determined by the spatial bandwidth of a real-time adaptive aberration-compensation system. The algorithm recognizes an object based on minimum distance in feature space. Each feature is associated with a corresponding scale size, which is correspondingly blurred by the OTF. Specific autocorrelation functions and probability distributions are assumed for the compensated aberrations (Gaussian) and the clutter strength (negative exponential and beta or Gaussian, respectively). The results presented assume an informationless distribution of targets in feature space and show the variation of performance with the number offeatures, the quality of compensation, and the strength of clutter.

Infrared presensitization photography (IRPP) is used to acquire images of the solar disk in the 1- to 2-μm regime. This is the first time IRPP has been employed at such short IR wavelengths, used against a thermal (instead of a laser) source, and applied to astronomy. The experiments demonstrate the feasibility of IRPP for solar imaging, but the image quality needs improvement for solar research

A series of experiments has been performed to determine the critical practical issues in high-density free-space optically interconnected systems. Some ofthese experiments have implemented switching fabrics by optically interconnecting 2-D arrays of symmetric seif-electro-optic effect devices (S-SEEDs). In the first, three 16 x 8 arrays of S-SEEDs, all operating as logic gates, were optically interconnected, and in the last experiment, afully interconnected switching fabric using six 32 x 32 S-SEED arrays was demonstrated. The practical realization of this technology represents a challenge to modern optomechanics because ofthe required optical resolution, mechanical precision, stability, and number of components involved. A comparison of these experimental systems shows that significant optical power loss may be incurred when theoretically "lossless" techniques are actually implemented, mainly because of the system complexity. The use of much simpler techniques is shown to dramatically decrease system assembly and alignment times and increase system stability, with similar overall loss. The tolerancing analysis used in these systems shows thatthe worst-case optical constraints result in mechanical tolerances in the micrometer to submicrometer range. The successful operation of these systems demonstrates the ability of relatively simple optical and mechanical techniques and materials to meet these tolerances.

This PDF file contains the editorial “Book Rvw: Optoelectronics and Lightwave Techology," by J. E. Midwinter and Y. L. Guo for OE Vol. 31 Issue 12

This PDF file contains the editorial “Book Rvw: Applied Optics and Optical Engineering, Volume XI," edited by R. R. Shannon and J. C. Wyant for OE Vol. 31 Issue 12