This paper describes an 0.75 meter aperture, Stabilized High-accuracy Optical Tracking System (SHOTS), two of which are being developed by Textron Systems Corporation, under contract to the Navy's Space and Naval Warfare Systems Center, San Diego (SPAWAR-SD). The SHOTS design is optimized to meet the requirements of the Navy's Theater Ballistic Missile Defense (TBMD) testing program being conducted at the Kauai Pacific Missile Range Facility (PMRF). The SHOTS utilizes a high-precision, GPS aided inertial navigation unit (INU) coupled with a 3-axis, rate gyro stabilized mount which allows precision pointing to be achieved on either land or sea-based platforms. The SHOTS mount control system architecture, acquisition, tracking and pointing (ATP) functionality and methodology which allows the system to meet the TBMD mission data collection requirements are discussed. High frame rate visible and MWIR sensors are incorporated into the system design to provide the capability of capturing short duration events, e.g., missile-target intercepts. These sensors along with the supporting high speed data acquisition, recording and control subsystems are described. Simulations of the SHOTS imaging performance in TBMD measurement scenarios are presented along with an example of the image improvement being achieved with post-processing image reconstruction algorithms.

Simulation development for EO Systems has progressed to new levels with the advent of COTS software tools such as Matlab/Simulink. These tools allow rapid reuse of simulation library routines. We have applied these tools to newly emerging Acquisition Tracking and Pointing (ATP) systems using many routines developed through a legacy to High Energy Laser programs such as AirBorne Laser, Space Based Laser, Tactical High Energy Laser, and The Air Force Research Laboratory projects associated with the Starfire Optical Range. The simulation architecture allows ease in testing various track algorithms under simulated scenes with the ability to rapidly vary system hardware parameters such as track sensor and track loop control systems. The atmospheric turbulence environment and associated optical distortion is simulated to high fidelity levels through the application of an atmospheric phase screen model to produce scintillation of the laser illuminator uplink. The particular ATP system simulated is a small transportable system for tracking satellites in a daytime environment and projects a low power laser and receives laser return from retro-reflector equipped satellites. The primary application of the ATP system (and therefore the simulation) is the determination of the illuminator beam profile, jitter, and scintillation of the low power laser at the satellite. The ATP system will serve as a test bed for satellite tracking in a high background during daytime. Of particular interest in this simulation is the ability to emulate the hardware modelogic within the simulation to test and refine system states and mode change decisions. Additionally, the simulation allows data from the hardware system tests to be imported into Matlab and to thereby drive the simulation or to be easily compared to simulation results.

This paper presents new simulation tools developed for the analysis of field and captive flight test data for imaging terminal homing missile systems. The analysis of field test data is required for both simulation validation and rapid design when development of a simulation model for each target and background scenario is cost or time prohibitive. In addition, the complex problem of resolving the inevitable performance differences between simulation and field test data will be addressed.

This paper presents recent design improvements for the Gray- Level Co-occurrence Matrix (GLCM) based Trackability Metric and the resulting performance enhancements. An in depth performance trade study for design modifications including a hot spot metric was performed. These enhancements to the original Trackability Metric should provide better state-of- the-art performance prediction and more accurate performance modeling for specific imaging autotracker designs. The results of the study and the final implementation of the Trackability Metric are presented.

This paper will describe a digital simulation model to assess the effectiveness of infrared imaging seeking missiles in presence of background and countermeasures. The model reproduces the infrared (IR) scene observed by the imaging seeker and generates the images at the detector output. The model replicates the image processing for automatic detection of potential targets in the seeker's field of view (FOV), and the target selection by the tracking algorithm. A model of a two-degree-of-freedom gyro stabilized seeker platform is also presented. The seeker platform is driven by the tracking algorithm to reduce the error between the seeker axis and the line of sight. The simulation also includes models for the missile aerodynamics in six degrees of freedom, the missile guidance and control system and the target and missile trajectories. As an illustration of the model, some results of simulation are shown for a target engaged by an infrared seeking missile using an imaging sensor.

The community of imaging missile system development is burdened by the custom build methodology for the implementation of Automatic Target Trackers (Autotrackers). Normally this is the most effective development path considering the high performance realtime requirements of an image processing system such as an Autotracker. However, with the progress made in modern and inexpensive compute platforms, it is now possible to achieve rapid insertion of the algorithm development platform with the sensor system, allowing refinement of algorithms and establishing known performance limits early in the development cycle. In past programs the Autotracker was often overlooked in importance, and was not given the financial resources for effective development within the system. In order to overcome this short coming, a cost effective Autotracker platform has been developed which can be readily inserted into most systems with minimum and sometimes no additional hardware or software required. We describe the hardware interface design of a realtime multi-mode Terminal-homing Missile Autotracker that has been optimized to utilize the vector functionality of the Macintosh resident G4 processor.

The use of synthetic or simulated target in electro-optical systems is an extremely useful where performance is dependent on operator skills. The synthetic targets can be inserted over the incoming background video or generated as a complete image with background and target within a high specification PC. This paper discusses both techniques.

The noise equivalent angle (NEA) is determined for a class of nose tracking algorithms that determine the leading edge by use of a partial body centroid algorithm. Analytical expressions for the track error and jitter variance are derived for silhouette and two types of intensity weighted centroid processing for a range of nose shapes determined by the parameter (alpha) . Graphical results are presented for triangular, parabolic and rectangle nose shapes. This paper only considers the effects of additive noise. Spatial quantization is not included.

In order to demonstrate lock-on-after-launch (LOAL) capability, imaging infrared missile systems of the future require the ability to autonomously identify and track targets of interest. A robust algorithm architecture must have the flexibility to accommodate the fluid system requirements that drive its design. This paper describes a method to autonomously acquire and track an extended range target through its entire flight scenario. A proven automatic target recognition (ATR) approach is used to detect and identify targets of interest, separating them into non-targets and clutter. The methodology uses a down selection strategy to nominate targets for terminal track. Once nominated, a Weighted Edge Tracker (WET) is employed. The tracker relies upon correlations of appropriately weighted edge directions from frame-to-frame images and reference templates. This combination of automatic target recognition and terminal tracking provides a sophisticated yet simple approach to many challenging long range tracking problems.

Conventional approach to object tracking through image sequence either use area correlation or use extraction of contrast edges and other features from the target image, taken as the target signature. However it is often difficult to reliably extract the target signature when the target and background appearance, as well as target/background polarity and contrast, change drastically over the course of the image sequence, and to maintain high accuracy of track in real-time. One of the major neurobiological discoveries in the last two decades is that the what processes, which determine the identity of an object, are segregated from the where processes, which determine the spatial location and motion of an object. Biological visual motion tracking does not require continuously detecting the target signature. It simply maintains the target signature throughout the image sequence via an adaptive process of the receptive fields of neurons. Formulated as a pure where process that maintains the Gabor representation of the target surface signature, we reduce the tracking process to the analytical computation of affine transformations of the surface signature through the image sequence. The great simplicity, high accuracy, and robustness of this analytical pure where process demonstrated the power of the biological computation strategy.

We present a technique for region tracking using a novel method of registration. This method uses the representation of the image with a tree form containing the topological and geometrical information in the image. The computation of the tree is achieved by using the so called Fast Level Set Transform (FLST) and owns the fine property of contrast invariance. We explain the main reasons why the shapes present in the image correspond to some nodes in the tree. We briefly describe the basic principles of the FLST. We then expose a registration algorithm, based on a vote procedure, that consists of finding correspondences between the nodes of the trees of two different images. We also porpose an implementation of this algorithm for a real time application and end with some numerical experiments.

The integration of multiple sensors for the purpose of forming a single integrated air picture has been intensely investigated in recent years. Assuming no sensor biases and minimal communication latencies, the ideal picture can be formed when all the sensor information is communicated to each network node. The state vector for each target at every node should be identical under this ideal condition. However, this is not the situation when sensor bias is considered since it has an adverse effect on the tracking performance by increasing the estimation error. A method to account for the location, measurement, and attitude biases of the sensors must be employed to provide more accurate state estimates of the target. This paper will present a method for estimating sensor measurement bias in a multi- target environment. The output of the bias estimation process will be employed to compensate the sensor measurements for the tracking of highly maneuvering aircraft. Utilizing the common tracks of multiple sensors, a comparison between compensated and uncompensated techniques will be provided through simulation.

Sensor data fusion has long been recognized as a means to improve target tracking. Common practice assumes that the sensors used an synchronous, i.e., they have identical data rate, measurements are taken at the same time, and have no communication delays between sensors platform and central processing center. Such assumptions are invalid in practice. This paper deals with removing such assumptions when considering the multi-sensor target tracking case. In particular, it assumes that the sensors used can have different data rates and communication delays between local and central platforms. A new target tracker using asynchronous sensors is proposed and derived in this paper. The performance of the filter is compared to the optical sequential filter using simulated targets.

Tracking algorithms typically rely on point-to-point correlation of sensor positional data. When target positional or identification data is obscured, missing, or noisy, tracking algorithm performance degrades and may lose track before a subsequent measurement is available. There are many cases when partial target detection results such as when a target travels behind an obstacle or is occluded by trees or industrial camouflage. In an effort to design tracking algorithms that can track through missing, occluded or data dropout, we seek to use a group tracker technique to solve the occluded target tracking problem. Two methodologies are employed to compensate for a partially observable target state and covariance: (1) a coasting individual targets (CIT) method and (2) a group-updated track (GUT) method. The coasting method is analogous to the tracking prediction equations and the novel group tracking update method recovers the unobservable target position state from the other members of the group.

We propose a new mobile station tracking algorithm based on the bootstrap filtering that uses the time difference of arrival (TDOA) measurements and the signal powers of the mobile station measured at several base stations. The proposed algorithm imposes nonlinear kinematic constraints on the state estimates without destabilizing the algorithm. Such constraints can be most naturally incorporated in the Bayesian bootstrap filtering framework (treating the constraints as perfect measurements without measurement noise). To verify the algorithm, simulation is performed and the result demonstrates that our algorithm gives better position estimate than the bootstrap filter without constraints or the constrained bootstrap filter using only one of TDOA measurement or signal power measurement.

This paper describes a new method for object movement estimation using sequences of images taken from a monocular camera. The method integrates a Kalman filter to estimate the three dimensional parameters of the optical system and a lineal projective model to determine 3D point coordinates projected on the retinal plane. The method works with at least three distinctive points in the image, and they are updated using correlation methods. The result is an estimation of the rotation and translation parameters between successive images within the sequence and yield to the 3D coordinates of the points selected for correspondence. The method is tested with synthetic images to evaluate its accuracy and later an interesting application in autonomous navigation is presented.

Novel sensor have been developed at the Army Research Laboratory (ARL) to provide continuous accurate angular measurements for spinning projectiles in free flight. These systems, which directly measure angular orientations, are distinct from angular rate sensor methodologies that require integration of these rates to estimate angular orientations. Also, many traditional rate sensors are expensive, voluminous and not well-suited to the high-g launch and high spin environment of many projectile-borne munitions. Recent advances in commercially-available magnetic sensors have yielded devices small enough, rugged enough, and/or sensitive enough to be used in body-fixed sensor constellations to make high-speed, high-resolution measurements of attitude and roll rate relative to earth's magnetic field. The addition of a complimentary sensor system measuring orientation relative to another distinct earth-fixed field of known orientation provides the information required to mathematically determine the absolute angular orientation of a spinning body within any desired navigation system, e.g., north, east, and vertical. Such a dual-field measurement system has been implemented utilizing a unique constellation of magnetoresistive sensors and ARL Solar Likeness Indicating Transducers (SLIT) to determine angular orientation with respect to the magnetic and solar fields respectively. The mathematical foundations of this dual-field sensor system will be summarized and flight experiments of the prototype systems will be discussed.

Laser satellite communication has become especially attractive in recent years. Because the laser beam width is narrow than in the RF or microwave range, the transmitted optical power may be significantly reduced. This leads to development of miniature communication systems with extremely low power consumption. On the other hand, the laser communication channel is very sensitive to vibrations of the optical platform. These vibrations cause angular noise in laser beam pointing, comparable to the laser beam width. As result, as significant portion of the optical power between transmitter and receiver is lost and the bit error rate is increased. Consequently, vibration noise control is a critical problem in laser satellite communication. The direction of the laser beam is corrected with a fast steering mirror (FSM). In this paper are presented two approaches for the FSM control. One is the feedback control that uses an LQG algorithm. The second is the direct feed- forward control when vibration noise is measured by three orthogonal accelerometers and drives directly the F SM. The performances of each approach are evaluated using MATLAB simulations.

The first step in creating an optical link between two LEO satellites is acquisition. In this process one of the satellites finds the maximal power of a received beam and locks on to it. This starts the tracking. In this paper we examine the time needed to finish the acquisition process and start tracking. The parameters included are the distribution function of satellite position, the size of the uncertainty area, the number of possible satellite positions, and the detection ability of a CCD. A model for the distribution function of position is given for two types of distribution: Gaussian and Uniform. Also considered the vibrations that come from internal systems of satellite, and from external sources. The characteristics of vibrations are considered and their influence on the scanning pattern that can deviate from the original path. A method of filtering the vibrations and compensating for them is suggested. The pointing system must be updated continuously from the star tracker with internal calculations of position, speed, velocity and vibrations characteristics. Examined also are several scanning methods: raster, spiral, Lissajo, Rose. Each method has its own possibilities and advantages, which are compared.

We propose the laser accelerometer for guidance and navigation as component for control systems. The linear laser has been sued to provide the sensitive light standing wave - baghron - in the proposed sensor. The custom optical components and the photodiode have been used to convert an interference pattern intensity signal to a reading of the acceleration. The main goal of the proposed efforts is to demonstrate the ability of the sensor to measure on the vehicle an acceleration of an irregular movement with respect to an inertial system in wide working range. Proposed sensor is being without any dead zone or hysteresis on output characteristics, it doesn't measure the gravity without motion and could be used in a large number of important applications. There are not in the accelerometer any parameters of the resonator attached/fixed with the object to be measured, which would be altering during the movement of the object. There is not any moving part or tensioned element in the proposed accelerometer defended with the US patent No 5,652,390. Date of patent: July 29, 1997. Class: 073-657.000.

The conventional proportional navigation studied before involves a guidance law that the commanded acceleration is regulated proportionally to the product of line-of-sight rate and closing speed. It is wildly used on the simple scenario of missile interception with small heading error. But on the dogfight scenario with high heading error, its performance is not so good as that with small heading error. In this article, an extended proportional navigation is introduced and studied. In this proposed guidance scheme, the commanded acceleration is regulated proportionally to the product of relative speed and line-of-sight rate between interceptor and its target, and it can be generalized to be applied in a different direction with a bias angle to the normal direction of line-of-sight. The relative speed is always greater than the closing speed and finally approaches to the closing speed till intercept. The exact and complete closed-form solutions are derived for both maneuvering and non-maneuvering targets. Some related important characteristics, such as capture capability and energy cost, are investigated and discussed. From the result, it is found that some improvement can be achieved under this new guidance scheme, as compared with the previous ones. Also, a typical example of target maneuver is introduced to describe the effect of target maneuver easily. It shows that the target maneuver will usually decrease the capture area and increase the energy cost for effective intercept of target.

This paper demonstrates an approach of combining the hyperbolic and triangulation techniques for better location of targets in multi-site sensor environments. The paper also addresses the various techniques and terms used to find target location in multi-site sensor environments. These include elliptic, hyperbolic, triangulation techniques and their combination. The hyperbolic technique is discussed in detail and simulation results are shown.

Applied Technology Associates' ARS-12 is the most sensitive inertial angular vibration sensor available in the market today. The sensing mechanism is based on magnetohydrodynamic (MHD) principles. This sensor has a bandwidth from 1-1000 Hz and a noise-equivalent angle of less than 35 nanoradians from 2-1000 Hz. The ARS-12 can measure inertial angular motions of less than 10 nanoradians at discrete frequencies. Their solid state design makes these sensors smaller and more rugged than any previous angular vibration sensor. In addition, the ARS-12 is essentially impervious to linear acceleration and angular cross-axis sensitivity is limited to incorrect physical alignment. The ARS-12 has recently undergone several design changes in order to survive the space environment. This new model, the ARS-12G, also has increased reliability and tighter performance specifications. The ARS-12G design, testing, and performance will be reviewed in this paper. Several ARS-12G sensor packages are currently being tested and space-qualified for Boeing(HSC) and Japan's space agency, NASDA.

Optical tracking of airborne targets typically involves initial acquisition by radar at long range and relatively high measurement error. The target is then passed to an electro-optical tracker with a WFOV and precision tracking begins. The problem can include a second handoff to a NFOV sensor for additional resolution. A stringent mission time line requires these handoffs to be executed quickly. In this paper, traditional approaches to these handoff problems are reviewed and a new solution is presented at the system level. The authors discuss problems uncovered in the integration and testing phase and show results from an extensive HWIL testing platform.