This PDF file contains the editorial “1997 in Review” for OE Vol. 37 Issue 02

This is the third offering of Optical Engineering special sections on sensor fusion, a field that has continued to flourish at a rapid pace during the past year.

The various decision fusion strategies postulated recently for target detection in the context of a three-sensor suite are analyzed in depth to delineate the domains of beneficial fusion and determine the extent of such benefits. This is similar in scope to earlier temporal fusion studies on suites with only two sensors, except that the present analysis is limited to fusion strategies involving only a single set of observations from the sensors. Instead, the added flexibility in the design of fusion strategies resulting from an increase in the number of sensors to three is defined and assessed. The analysis covers two basic categories of fusion strategies. The first is a single-stage fusion wherein the decision outputs of all the three sensor subsystems are fused simultaneously using one of four different strategies, namely, AND logic, simple majority logic, firm-decisions-only majority logic, and a no-firm-contradiction logic. The second is a two-stage fusion strategy, wherein either an AND or an OR logic is used at the first stage combining the decisions of two of the sensor subsystems, followed by a similar logic choice combining the fused decision from the first level with the decision from the third sensor subsystem, resulting once again in four possible alternative strategies. First, each of the strategies is compared to the single-sensor performance to delineate the corresponding domain and extent of the fusion benefits, using for illustrative convenience the special case of a scenario wherein all the three sensors in the suite have identical expected performance characteristics. Next, a detailed comparative assessment of these strategies is made under the same assumption of identical sensor characteristics. The set of eight fusion strategies under the three-sensor suite is then compared with the two simple Boolean (AND logic and OR logic) strategies under the two-sensor suite to delineate the benefits domain of adding the third sensor to the suite.

A hybrid decision level architecture for a road collision risks avoidance system is presented. The goal of the decision level is to classify the behavior of the vehicles observed by a smart system or vehicle. The knowledge of vehicle behavior enables the best management of the smart system resources. The association of a model to each observed vehicle mainly enables the limitation of inference and of the set of actions to be activated; thus the interactions between system levels can be more intelligent. The decision level of this architecture is composed of a neural classifier, which is associated to a numerical classifier. Each of these classifiers provides decisions that are expressed within the framework of fuzzy theory. An optimal fusion policy is reached using the functional neural network tool.

A new fusion architecture is presented and some preliminary results on field data are shown. The architecture consists of a new decision level fusion algorithm, the piecewise level fusion algorithm (PLFA), integrated with an expert system based user assistant. The PLFA technique extends a known two class fusion decision level detection technique to a multiple class fusion classifier technique. The expert system assistant calculates the parameters necessary to automatically adapt the PLFAtothe required performance. An interactive integration of the PLFA module with the expert system module allows for effective search for fusion parameters while shielding the user from the mathematical complexity of the fusion procedure. This integration enables the detection of deadlocks and the generation of guidance to the user in parameter selection. The overall architecture and processing required to develop the current system from a man-in-the-loop system to a fully automated fusion system are also described.

A case study in which the Dempster-Shafer evidential reasoning theory is applied to a naval identity and attribute fusion problem is described. Two identity information fusion architectures are discussed: the first concerns the fusion of identity declarations where the sources are expected to provide only useful and complete results such as an identity declaration, and the second architecture addresses the fusion of incomplete attribute information coming from various sources. The emphasis is put on the second architecture. The use of fuzzy logic for proposition generation is described as well as techniques for managing the number of propositions generated.

The problem of orthogonalization of multiple satellite images, focusing at the same field of view, is studied in this paper for constant false alarm rate (CFAR) point target detection. Information fusion techniques are incorporated. The Karhunen-Loéve (K-L) orthogonalization method offers the best CFAR detection performance. However, it requires central fusion. It is shown that a version of the Gram-Schmidt (G-S) orthogonalization method, which preprocesses data in a pipeline form, offers a comparable CFAR detection to that of the K-L method. Point target CFAR detection is carried out by employing various fusion approaches on the orthogonal data. Sensor level fusion with quality information is shown to be preferable when the proposed sequential G-S orthogonalization is applied. The proposed CFAr approach is applied to dissimilar sensors and avoids overloading the communication channel transmitting only in the case of target detection. Trade-off studies and experimental results on real and simulated data are presented.

We present a flexible hybrid decision scheme for decentralized detection under communication constraints. In this scheme, local sensors send a binary (hard) decision to the fusion center when the local sensors have a relatively high confidence in the decision, otherwise a perfect version of the local likelihood ratio (LLR) is sent. In practice, a finely quantized version of the LLR is sent. The degree of confidence at which this switch is made is determined by the specified communication constraint. The fusion center makes a final decision based on the information received from local sensors. By employing the person-by-person optimization methodology, we develop the local decision rules and the fusion rule. Owing to the associated computational difficulty, we propose a simpler procedure based on the class of Ali-Silvey distance measures to obtain the local decision rules. A numerical example is also presented for illustration.

In practice, multisensor systems use dissimilar sensors that may have different data rates. Such sensors may also have inherent delays due to multitasking as well as communication delays between the sensor platform and a remote central processing site. Track fusion algorithms are presented that are valid for asynchronous sensors (the sensors have different data rates and different delays) as well as synchronous sensors (all of the sensors take measurements at the same time and the same rate with no delays). The asynchronous track fusion problem is formulated and solved first. Then the synchronous track fusion problem is obtained as a special case of the asynchronous one. Finally, using simulated target tracks, the performance of the asynchronous track fusion (ASTF) algorithm is analyzed and compared to an existing track fusion algorithm. Different sensor data rates and communication delays are used in the simulations. It is found that the ASTF algorithm outperforms its counterpart and is able to handle relatively large communication delays. The results presented set the foundation for deriving optimal track fusion algorithms when taking into account realistic constraints such as sensors with different data rates and different communication and/or processing delays. The results can also be used as a benchmark to evaluate existing suboptimal track fusion algorithms.

The results of a quantitative comparison of two architectural options in developing a multisensor data fusion system are presented. One option is the contact-level or centralized architecture: a single track file is maintained and updated using raw sensor measurements. The second option is the autonomous or track-level sensor fusion architecture: each sensor maintains its own track file. The sensor tracks are then transmitted to a central processor responsible for fusing this data to form a master track file. The emphasis for this study is given to measuring the quality of the fused product and the bandwidth of the communications required between the sensors and the fusion center. The evaluation is conducted with the Concept Analysis and Simulation Environment for Automatic Target Tracking and Identification testbed. This testbed provides the algorithm-level test and replacement capability required to conduct this kind of performance study.

In a multiplatform multisensor system, it is crucial to develop a practical algorithm under computation and communication constraints. A practical data association algorithm is presented. A decoupling probabilistic data association (DPDA) algorithm is proposed. The new algorithm decouples the joint probabilistic data association (JPDA) algorithm into separate PDA algorithms in a simple way because the complete decoupling PDA algorithm is too complex and loses its meaning for realtime use when there are more than three correlated targets. The algorithm is then extended to multiplatform multisensor tracking, where it is combined with a distributed algorithm. To evaluate a data association algorithm, the correct average probability (CAP) is used. The CAP is defined as the average probability of a true target measurement. Monte Carlo simulations demonstrate the effectiveness of the algorithm.

A distributed multirate interacting multiple model (DMRIMM) tracking algorithm is developed. The algorithm is based on a reformulation of the interacting multiple model (IMM) filter under the assumption that each model operates at an update rate proportional to the model's assumed dynamics. A multirate model is derived based on the geometrical interpretation of a discrete wavelet transform. A wavelet transform is used to generate equivalent multirate measurements, which exhibit the additional property of lower equivalent measurement noise for low-rate data. Using this filtering approach, the MRIMM algorithm significantly outperforms a full-rate IMM filter when no maneuvers occur and moderately outperforms the IMM filter when maneuvers occur. Based on the MRIMM algorithm, a distributed version DMRIMM algorithm for multiplatform multisensor tracking is proposed. The MRIMM algorithm is first employed to perform each local/platform estimation. A global filter is then constructed to perform a fusion of MRIMM estimations. The advantage of performance improvement is demonstrated through Monte Carlo simulations.

A centralized track fusion algorithm that incorporates crosscorrelation between tracks originating from multiple platforms at remote sites is described. Remote stations transmit the information contained in these track updates to a central station where the tracks are kinematically fused to create composite tracks. it has been established that the probability distribution of correct track-to-track association can be improved if the cross-covariance matrix between the candidate tracks for fusion is positive. Necessary and sufficient conditions for positivity were derived and the steady state solution of the cross-covariance matrix was obtained in terms of the parameters of the candidate tracks to be fused. However, system-theoretic implications of these constraints were unclear. We obtain the steady state solution of the cross-covariance matrix in terms of a line integral. It is shown that evaluation of this integral involves inversion of an asymmetric matrix. For fusion of tracks created by similar sensors, this matrix is reduced to that of the Schur matrix, which arises in the analysis of steady state stability of the tracker associated with each sensor. However, for fusion of tracks created by dissimilar sensors, the structure of the matrix to be inverted is complicated and can not readily be partitioned in terms of the Schur matrices associated with each tracker. An efficient algorithm for the inversion of this matrix is also presented.

An approach is presented to the long range automatic detection of vehicles, using multisensor image sequences. The method is tested on a database of multispectral image sequences, acquired under diverse operational conditions. The approach consists of two parts. The first part uses a semisupervised approach, based on texture parameters, for detecting stationary targets. For each type of sensor one learning image is chosen. Texture parameters are calculated at each pixel of the learning images and are combined using logistic regression into a value that represents the conditional probability that the pixel belongs to a target given the texture parameters. The actual detection algorithm applies the same combination to the texture features calculated on the remainder of the database (test images). When the results of this feature-level fusion are stored as an image, the local maxima correspond to likely target positions. These feature-level-fused images are calculated for each sensor. in a sensor fusion step, the results obtained per sensor are then combined again. Region growing around the local maxima is then used to detect the targets. The second part of the algorithm searches for moving targets. To detect moving vehicles, any motion of the sensor must be detected first. If sensor motion is detected, it is estimated using a Markov random field model. Available prior knowledge about the sensor motion is used to simplify the motion estimation. The estimate is used to warp past images onto the current image in a temporal fusion approach and moving targets are detected by thresholding the difference between the original and warped images. Decision level fusion combines the results from both parts of the algorithm.

An image fusion procedure based on fuzzy set theory that can be used to classify different image components and also to preserve and even emphasize the internal contrast is presented. Membership functions are utilized to quantitatively define the relationships between different image classes, as well as the systematic and stochastic measurement errors, in terms of pixel values. For each modality, a possibility measure is applied to determine the degree to which each pixel belongs to various image classes. These possibility measures are sent to an image fusion center, where the image components are classified and their internal contrast restored and augmented. The methodology is practically applicable even in severely noisy environments. Results generated by the proposed method illustrate its capabilities in classifying and preserving internal image details.

Image fusion is a difficult task since the volume of data is quite large and the target is usually not completely represented in any single channel. Results from a spiral image fusion technique based on the pulse-coupled neural network to fuse channels into a set of single channel complex images are presented. These new images can then be filtered by a Fourier filter to find the target. The example used to demonstrate this technique uses images from a 30 channel multispectral sensor viewing a single scene.

In a multiple sensor system, the sensor Sj , j=1,2, . . . ,N, outputs Y( j )?R in response to input X?[0,1], according to an unknown probability distribution PY(j )|X . The problem is to estimate a fusion function f:RN?[0,1], based on a training sample, such that the expected square error is minimized over a family of functions F that constitutes a finite-dimensional vector space. The function f* that exactly minimizes the expected error cannot be computed since the underlying distributions are unknown, and only an approximation f to f* is feasible. We estimate the sample size sufficiently to ensure that an estimator f that minimizes the empirical square error provides a close approximation to f* with a high probability. The advantages of vector space methods are twofold: (1) the sample size estimate is a simple function of the dimensionality of F and (2) the estimate f can be easily computed by the well-known least square methods in polynomial time. The results are applicable to the classical potential function method as well as to a recently proposed class of sigmoidal feedforward neural networks.

Many sensor fusion systems combine redundant inputs to increase information reliability. In spite of this, few studies show how to choose redundant sensors for these systems. We find sensor configurations that minimize system cost while ensuring system dependability. Dependability is the generic term for system reliability and availability. Given many types of sensors, all fulfilling system operational requirements, but with different dependability and per item cost, heuristic search methods are used to find minimum cost configurations. Our main contributions are deriving the optimization problem, showing the search can be limited to a multidimensional surface, deriving a fitness function, and providing an efficient algorithm for computing dependability bounds. Two heuristics, genetic algorithms and simulated annealing, are proposed as methods. Experimental results show cost savings of up to 20% compared to systems with only one component type.

An enhanced architecture for a sensor measurement scheduler and a dynamic sensor scheduling algorithm called the on-line, greedy, urgency-driven, preemptive scheduling algorithm (OGUPSA) are presented. The premise is that the function of sensor management can be partitioned into the two tasks of information management, essentially an information to measurement mapping, and a sensor scheduler that takes the measurement requests along with their priorities and optimally maps them to a set of sensors. OGUPSA was developed using the three main scheduling policies of most-urgent-first to pick a task, earliest- completed-first to select a sensor, and least-versatile-first to resolve ties. By successive application of these policies. OGUPSA dynamically allocates, schedules, and distributes a set of measurement tasks from an information manager among a set of sensors. OGUPSA can detect the failure of a measurement task to meet a deadline and improves the dynamic load balance among all sensors while being a polynomial time algorithm. One of the key components of OGUPSA is the information in the applicable sensor table. This table is the mechanism that is used to assign requested tasks to specific sensors.

We consider the problem of sensor alignment for multisensor data fusion. A least squares (LS) sensor alignment algorithm is formulated using the earth-centered earth-fixed (ECEF) coordinate system. Perturbation analysis is carried for the ECEF-LS algorithm in the presence of sensor elevation measurement errors. A modified generalized least squares (MGLS) alignment approach is proposed that is applicable to 2-D sensor fusion systems and a computationally stable algorithm is developed to obtain the MGLS solution. Computer simulations are used to show the effectiveness of the proposed alignment algorithm.

We introduce and demonstrate the use of wavelet-filter-based processing in a joint transform correlator for achieving high-quality target discrimination. The inherent frequency localization characteristic of these filters offers high discrimination in correlation for lookalike targets. Results obtained from simulation and experimental studies validate this fact. Wavelet filters are generated and applied to the joint power spectrum through fast digital processing and are displayed onto a spatial light modulator. The correlation results are analyzed for target discrimination. The hybrid experimental setup combines the speed of optical processing with the flexibility of digital processing to realize such a technique in real time.

Nonimaging edge-ray concentrators may be uniquely well suited to a number of laser fiber optic contact surgical procedures where sizable increases in irradiance are required at maximum collection efficiency, and light must be emitted uniformly over a wide angular range. Compared to earlier fiber optic devices developed for these tasks, the delivered irradiance from nonimaging designs can be more than double, at maximum radiative efficiency, while producing a Lambertian absorber, i.e., uniform emission over the entire hemisphere. Because edge-ray designs appear to have been overlooked in the development of laser fiber optic devices, we present a couple of candidate designs and evaluate attainable flux concentration and radiative efficiency.

Polarization characteristics of Rayleigh backscattered light in erbium-doped optical fiber amplifiers are investigated by using 1.54-?m signal light pumped by 1.48-?m light. The Rayleigh backscattered equivalent reflectance for the signal light was ? 49.9 dB in a 30-m-length erbium-doped fiber and was increased and was amplified with the pump light. Linear polarization in the input signal light was almost preserved in the Rayleigh backscattered light in the absence of pumping. The degree of polarization was degraded from 88% to 52% with an increase in the pump power.

A synthetic hologram depicting a natural scene faces the difficulty of recording a tremendous number of pictures. This paper presents an idea to solve this problem. The scheme needs only few pictures, but can give satisfactory visual effects by a simple and fast algorithm. It is based on the matching process of stereo vision. Analysis and experimental results illustrate the feasibility of the approach.

This paper proposes a color image coding method based on recurrent iterated function systems (RIFSs). To encode a set of multispectral images, we apply a RiFS to multiset data consisting of three images. in the proposed method, the mappings not only between blocks within an individual spectral image but also between blocks of different spectral images are performed with contraction constraint. Simulation results show that the fractal coding based on the RiFS is useful for encoding concurrently a set of images by describing the similarity existing between a pair of images. In addition, the proposed color coding method can be applied to subband images and moving image sequences consisting of a set of images having similar gray patterns.

An efficient technique for restoring the original from a blurred image using the concept of fringe-adjusted joint transform correlation is proposed. This technique employs Fourier plane apodization using the reference image power spectrum to deblur the distorted image. The performance of the proposed technique has been enhanced significantly by employing Fourier plane image subtraction, especially for noisy input scenes involving multiple objects. The image subtraction technique eliminates the unwanted zero-order term and cross-correlation terms produced by similar input scene objects while alleviating the effect of noise that may be present in the unknown input scene. Computer simulation results using noise-free and noise-corrupted input scenes are presented to verify the performance of the proposed technique.

We investigate the performance of an image encryption technique that uses random-phase encoding in both the input plane and the Fourier plane, using partial information of the encrypted image. We first investigate the phase-only information of the encrypted data for decryption. A binary version of the phase-only information is also considered for decryption. Binary images are well suited for optical display and practical implementation. Using partial information of the encrypted image, a reconstructed complex image is generated, which is used for decryption. Tests are performed for both gray-scale and binary images. We show that the phase information of the encrypted image is very important in the reconstruction of the decrypted image. Computer simulations show that for the images tested here, binarization of the encrypted image can recover the original image with low mean squared error.

Traditional image segmentation techniques typically divide an image into separate regions based on gray-scale characteristics. Most real-world image-segmentation problems, however, require some subsequent shape-based processing to yield acceptable results. Unfortunately, choosing an appropriate sequence of image-processing operators (a process) for this purpose can be a time-consuming, tedious procedure that requires considerable image-processing expertise. We describe a semiautomatic paradigm for selecting shape-based operations for an image-analysis process. Desired shape information for image regions is provided by the user in the form of easily specified cues. The cues are then automatically interpreted to select suitable imageprocessing operators and operator parameters; the operators can be morphological, topological, and image-manipulation functions. The paradigm, hence, enables easy prototyping of image-analysis processes for different problems. The user is not required to be an image-processing expert to apply this strategy—he or she need only be able to specify the desired shape properties of the regions in the image. We demonstrate our approach for both 2-D and 3-D image analysis problems.

The goal of this research is to find signature strength metrics that correspond closely to the distinctness of a target as perceived by humans, especially for camouflaged targets. We consider metrics that attempt to measure the strength of three perceptual cues: contrast, texture differences, and boundary strength. The contrast cue is measured with first-order metrics, differences in texture are measured with a second-order metric, and boundary strength is measured by computing contrast along the target-background boundary. We discuss a psychophysical experiment designed to generate quantitative measurements of perceived target distinctness for comparison with the target signature strength metrics. This experiment involves paired comparisons of image stimuli, each containing a single random target pattern embedded at a known location in a random background pattern. The data from the psychophysical experiment are compared with computed values of the target signature strength metrics, and a second-order image texture metric was found to exhibit the strongest correlation with the human data.

The imaging characteristics of a birefringent lens sandwiched between two polarizers is studied analytically. If the optic axis of the birefringent crystal is perpendicular to the lens axis, the system behaves as a bifocal lens for proper orientations of the polarizers. It is shown that the proposed system can be adapted either for apodization or for superresolution simply by rotating either of the two polarizers. It has greater depth of focus than conventional systems, as is evident from the results obtained. Expressions are obtained for the point spread function of such a system. Some specific cases are illustrated graphically.

We first review the mathematical physics of quantum decoherence in generic single-qubit devices, including environmental interactions. Next, we formulate the persistence probability for a qubit device as the probability of measuring the qubits in their unperturbed state without the decoherence arising from environmental interactions. A decoherence time can be obtained from the persistence probability. Drawing on recent work of Garg, and also Palma, Suomine, and Ekert, we apply the persistence probability formalism to a generic single-qubit device coupled to a thermal environment, and also apply it to a trapped-ion quantum register coupled to its ion vibrational modes.

A field rotator dedicated to compensate for the celestial field rotation has been designed and built for the Grand Interferometre a 2 Telescopes (GI2T) by the Laboratoire d'Astronomie Spatiale (LAS). In each interferometric arm, we introduce a three-reflection optics rotatable around the axis of the optical train. Since the visibility of the interference fringe pattern is mostly sensitive to the differential instrumental polarization, this device must fulfill not only optomechanical requirements but also polarization constraints. For this purpose, three metallic reflections at small incidences are preferred to classical K-prisms (such as Dove or Kdenig-Abbe prisms) and coatings optimization leads to protected silver mirrors.

A noncontact optical heterodyne technique was implemented for in situ measurement of crystal growth rate from solutions. A two- frequency Zeeman laser with a total output power of 1 mW (each polarized component is 0.5 mW) was used at 633 nm and is composed of two frequencies v1 and v2 that are collinear but polarized orthogonally to each other and separated by precisely 250 kHz. The sensitivity of the system was found to be 0.77 nm. This value is larger than the theoretical resolution limit of 0.065 nm because of the noise associated with the vibrations. The critical factors of the heterodyne technique for the monitoring of solution crystal growth have been studied. The growth rate of L-arginine phosphate crystal was 6.15±0.13 nm/s (0.53 mm/day) at the solution temperature of 28.2°C and seed crystal temperature of 26.0°C. This is the first experiment in which the heterodyne technique has been applied for real-time in situ measurement of crystal growth rate.

Since current fringe analysis techniques used in electronic speckle pattern interferometry (ESPI) yield a phase map modulo 2t, phase unwrapping is the final step of any data evaluation process. In this paper the performance of a recently published unwrapping algorithm is investigated. The algorithm is based on a least-squares minimization technique solvable in an iterative manner by the discrete cosine transform. Phase inconsistencies are handled by excluding invalid pixels from the unwrapping process through the assignment of zero-valued weights. The study is carried out using computer simulated ESPI phase maps, an approach which allows us to generate noise-free phase fields that are used to calculate phase deviations as a function of the iteration number.

We describe a method to determine the anisotropic parameters of liquid crystal televisions (LCTVs) formed by twisted nematic liquid crystals, as a function of the wavelength over a broad spectral range. One of them is the twist angle ? and the second, ?, is related to the birefringence of the liquid crystal. The method is based on the concept of the channeled spectrum and applies whenever the user has no information about the structure of the cell and the molecular director. It is shown that the recording of three channeled spectra is sufficient to determine the twist angle and the retardation properties of the cell. The method is illustrated with the experimental characterization of red, green, and blue LCTVs from a video projector.

We present a new autofocusing method as a part of a general project to automatize a transmission microscope for inspection of cytological tissue. These images present several cell superpositions that lead to the appearance of false edges under defocus conditions. This means that some autofocusing algorithms, based on derivatives, may not work properly. The method that we propose allows us to solve this problem. The paper is divided in two parts. In the first we present a new focusing criterion, the dynamic focus criterion (DFC). It is computed by comparing (point to point) the gray-level values of two images instead of comparing the energy of the first or second derivatives. This allows us to avoid certain bad effects on the shape of the criterion function, which are due to the structure of the cytological images and the transmission of light. In the second part we propose an optimal search algorithm to find the best-focused image using the DFC. Optimality is given in terms of the number of lens movements. The algorithm considers not only the unimodal property, as the classical Fibonacci search does, but also the symmetry of the criterion function. Finally, some results on its performance are given.

A moire method for dynamic in-plane displacement measurement is used with the purpose of studying carbon fiber/epoxy composite specimens under tensile dynamic load. The displacement field is measured with a random noise level of less than 0.3 ?m and by using a high-speed camera, a framing speed of 1 million frames/s is achieved. Cracks propagating at speeds of 2000 m/s have been observed. To evaluate strain accurately from the displacement fields, a calibration for distortion in the high-speed image converter camera must be made. A method to measure the variable magnifications in the different frame areas of the camera back is described.

Multilayer perceptrons (MLPs) have been widely applied to pattern recognition. It is found that when the data has a multimodal distribution, a standard MLP is hard to train and a valid neural network classifier is difficult to obtain. We propose a two-phase learning modular (TLM) neural network architecture to tackle this problem. The basic idea is to transform the multimodal distribution into a known and more learn- able distribution and then use a standard MLP to classify the new data. The transformation is accomplished by decomposing the input feature space into several subspaces and training several MLPs with samples in the subsets. We verified this idea with a two-class classification example and applied the TLM to the inverse synthetic aperture radar (ISAR) automatic target recognition (ATR), and compared its performance with that of the MLP. Experiments show that the MLP is difficult to train. Its performance depends strongly on the number of training samples as well as the architecture parameters. On the other hand, the TLM is much easier to train and yields better performance. In addition, the TLM's performance is more robust.

Polarization dependence of photopolymer-based volume holograms for substrate-guided-wave optical interconnects with surface normal configuration is investigated. Energy optimization and the tradeoff between energy equalization and polarization insensitivity of one-to- many cascaded interconnects are discussed. DuPont photopolymer film HRF 600X001-20 was employed in our experiment. A volume hologram with diffraction efficiency of 91.5%± 1.5% is achieved for different linearly polarized optical waves. A 1-to-9 fanout device with ± 5% energy fluctuation for a randomly polarized input optical wave at 632.8 nm has been fabricated for evaluating the trade-off under p and s waves. As the Mylar substrate of the photopolymer films has strong effects in changing incident linearly polarized light into elliptically polarized light, the energy uniformity of a practical device can be improved significantly. The developed theory is applicable to any volume hologram.

We report progress in the fabrication of very low areal density (<5 kg/m2) optical mirrors using space-flight-qualified graphite-fiberreinforced cyanate ester composite materials. Previous attempts have been thwarted by fiber print-through at the surface. We find that the problem can be successfully overcome if the composite is processed to leave a very thin layer of resin at the surface. Results thus far include replicas with surface microroughness ,1 nm rms, midfrequency ripples <3 nm rms, areal density 2 kg/m2 at 42-cm aperture, and freedom from print-through after vacuum drying and ion milling. The process is being extended to the fabrication of very lightweight meter-class optics for space UV astronomy applications, x-ray optics, and other ground- and space-based applications.

Perfluorinated pellicles are screened for thickness uniformity, absence of birefringence, photolytic stability under prolonged UV irradiation, and survival strength under pulsed (0.5-ns) high-peak-power laser irradiation. Results prove them to be extraordinarily robust against laser damage, photolytically stable, and better in transmitted wavefront quality than can be measured by state-of-the-art interferometry.

A method that employs an integrating sphere and a singlewavelength (1.6-?m) pyrometer to measure the spectral effective emis- sivities of YSZ thermal barrier coatings in the temperature range (~650 to 1050°C) is described. The effective spectral emissivities for two YSZ samples are measured as a function of the heat cycle duration. An increase in the emissivity of the YSZ samples is observed after the first heat cycle. Such an increase is attributed to the diffusion of aluminum and aluminum oxide from the bond-coat into the YSZ layer. The average effective spectral emissivity measured for YSZ after the first heat cycle was 0.729± 0.013. © 1998 Society of Photo-Optical Instrumentation Engineers.

In high-quality barcode readers, specular reflection from shiny barcodes is blocked by using a polarized scan laser and a crossed polarizer in front of the detector. When complex scanning geometries are required, the polarization properties of the mirrors in the system can become a limiting factor in system performance. Polarization raytracing allows systems such as barcode readers, LIDAR systems, and other polarization-critical systems to be accurately characterized. Polarization raytracing often requires the use of a local, ray-based coordinate system for expressing rays’ polarization states, yet the choice of coordinate system can have important implications on system analysis. An example is presented in which specular reflection is controlled in a barcode reader by using reflection-enhanced coatings on only one of the four sets of the mirrors in the system. The coordinate system used to express rays’ polarization states in the example system provides useful lessons for other systems. The other analytical methods used in this example can be applied to a variety of scanning systems.

The possibility of determining interface roughness and cross-correlation statistics of the two interfaces of a thin film from angle-resolved light scattering data at three wavelengths is investigated. It is shown that angle-resolved light scattering measurements at three wavelengths are not sufficient to determine the three power spectral density functions describing the thin film roughness. An attempt to combine reflectance and transmittance scattering to determine the roughness of a thin film on a transparent substrate appears to work and provides encouraging results.

A simple procedure characterizing light attenuation by transmission (based on the diffusion approximation of the transfer theory, the asymmetry factor, and the effective refractive index of the scattering medium) is compared with the coherent approaches based on (1) published Monte Carlo simulations, (2) scattering from a system of closely packed spherical particles with interparticle correlation characterized by the Percus-Yevick approximation, and (3) experimental data. Use of the procedure in the analysis of phase separation in polymer blends is presented. It is shown that the results obtained by this procedure are in reasonable accord with theoretical and experimental data so far published for systems with a volume fraction of minor phase less than ~0.2 and without any specific interaction in the system.

Radiometric relations are developed for a pixel-based optical system. Expressions for the contrast of both full- and subpixel objects are developed. Detector characteristics are incorporated. Object and background orientations are addressed. Both exact and approximate implementations of the modulation transfer function are provided.

An improved model of wavelength division multiplexing (WDM) by superimposed gratings in planar waveguides has been developed. Based on this theory, principal design rules of N-channel WDM are established and a fanout capacity of 100 is estimated. Finally, fourchannel WDM at 840 nm has been demonstrated on a Ge:SiO2 /SiO2 /Si planar waveguide, which was hydrogenated to enhance its photosensitivity. The superimposed gratings were written using a narrow linewidth KrF excimer laser in an interferometric setup used in inscribing fiber Bragg gratings.