Due to the large dimensions of third generation synchrotron radiation sources, no optical element can be placed closer than 25-30 meters away from the source. As a result, some of the standard x-ray optics layouts cannot be used without accepting a severe loss of performance. In an effort to overcome those limitations we decided to get away from the 1 to 1 toroidal focusing optics that is routinely used in synchrotron x-ray applications and, instead, to base out beam line optical design on a bendable conical x-ray mirror with unequal vertical and horizontal focal lengths. This raised some questions regarding both the optical performance of such mirror and the existence of the technology to fabricate it. Both questions were in principle answered positively. This paper discusses detail of the optical design and performance analysis and presents the results obtained while developing a fabrication process capable of producing conical x-ray optic.

Most asteroids subtend extremely small angles in the sky so it is impossible to directly discern either surface reflectivity, shape variations, or the spin axis using existing ground-based telescope systems. A super-resolution inversion technique relates the lightcurves of a rotating asteroid, which is represented as a polyhedron with planar surface facets, to the unknown radial vertex distances. Any number of lightcurves at different aspect angles can be incorporated into the formalism. Generalizations to delay-Doppler echo radar imaging are also discussed, as well as some of the issues associated with these inversion methods.

Matrix lightcurve inversion (MLI) is an indirect imaging technique that can be used to infer surface albedo distributions, pole orientations, and 3D shapes for planets, moons, and asteroids from rotational photometric lightcurves. Also, MLI can be used to map the brightness distributions of magnetically active spotted stars. Pluto represents an almost ideal application for MLI, because it is a spherical body with a known pole orientation and it is always in opposition to the Sun as observed from Earth. There is evidence that an extensive covering of methane or nitrogen frost sublimates as Pluto approaches perihelion, exposing a dark layer of photolyzed methane or nitrogen in an uneven pattern, and then freezes out again as Pluto passes perihelion, once again covering the planet. Assuming that this hypothesis is correct, we present a 'snapshot' of Pluto as it would have appeared during the years 1980 to 1986. Recent Hubble Space Telescope (HST) images of Pluto, although not as detailed as the image presented here, tend to confirm our snapshot of Pluto.

Bandpass filters using absorptive materials were developed to filter the incandescent light sources commonly found in crewstations to provide lighting compatible with Night Vision Imaging Systems (NVIS). A discussion on how NVIS operate provides insight into the rationale behind the military specification for NVIS compatible lighting; the radiance and color of NVIS compatible lighting is also examined. To achieve the desired results, a filter designer will find a balance between four different interdependent parameters. Finally, recent developments in NVIS filter technology are examined to determine current trends.

High resolution electronic imaging is certain to become the dominant method for recording microscope images (photomicrographs) in industrial laboratories. The rate of transition from film recording to electronic recording depends upon cost and quality. Cost, with little regard for image quality, has already caused a significant shift away from film recording. Recent technology developments appear to have eliminated the quality limitations of electronic imaging for industrial microscopy. These quality limitations are primarily related to the inability to match tradition 4 by 5 photomicrographs in field size and resolution. Analysis and limited experimental results indicate that 1024 by 1280 pixel imaging should be able to achieve the objective of matching 4 by 5 photomicrographs.

Biomedical diagnostic systems perform as combinations of multiple components, each introducing an experimental uncertainty. The present study examines the uncertainty in spectrophotometric measurements of absorbance A for different models of detector response. Modeling this response provides both a quantitative estimate of the minimum relative concentration error (RCE) and limits on the operating range for which the RCE falls within specified limits. For the constant uncertainty model the operating range corresponding to RCE no greater than twice the minimum value ranges from A approximately 0.07 to A approximately 0.79. A photomultiplier-like response results in the limits 0.2 < A < 2.3. A Beer's Law model predicts that RCE contributions from sample concentration uncertainty decrease monotonically with increasing absorbance. Assay chemistry factors dominate at low A values making high precision difficult at low A.

A photorefractive dynamic Schlieren (PDS) system is designed to monitor the photothermal damage-threshold of ZrSiO₂. The PDS is calibrated by observing the fringe movement in a Twyman-Green interferometer (TGI). While the phase changes due to optical path difference (OPD) of about (gamma) /10 are detectable by TGI, the PDS is capable of revealing intensity changes due to OPD of less than (gamma) /20. In this experiment a Nd:YAG laser is the source for photothermal phase change while a HeCd provides the Schlieren field.

Radiation sensitive films provide a dual purpose for monitoring the exposures received by people working with x-ray machines and radionuclides. In addition to giving quantitative estimates of radiation dose, radiological films yield images that often enable the radiation physicist to assess the physical conditions that existed during exposure. Optically stimulated luminescence (OSL) dosimetry represents a new radiation measurement technology that promises to provide a radiological image useful for radiation protection. OSL operates on the principal that certain crystals exposed to ionizing radiation can be made to luminesce following stimulation with selected frequencies of light. The amount of luminescence is directly proportional to radiation dose. OSL eliminates many of radiographic film's disadvantage such as sensitivity to heat and humidity, the need for chemical development and limited dose measurement range. Presented are results of using sapphire powders embedded in thin acrylic sheets and employing an OSL method in which stimulation occurs under cryogenic conditions. Landauer intends to use this method in a large-scale laboratory environment.

Radioactive scintillators have been used to calibrate photodetectors for many years. When these photodetectors are photon counters, an unusual calibration error occurs. Differences in the throughput or quantum efficiency of detection systems normally produce corresponding differences in the amount of light measured by the detection systems. But scintillators can produce pulses composed of several photons. When these sources are used to calibrate photon counting detection systems, differences in throughput can cause differences in pulse intensity rather than differences in pulse rate. Since photon counters are actually pulse rate detectors, differences in the pulse rate produced by photon counters may not accurately reflect differences in the throughput of these detection systems. This paper will present data and a model illustrating this problem, as well as a solution for the problem. The paper will also show how this effect can be used for absolute measurement of the efficiency of a detection system.

A method of evaluating the uniformity of an illumination pattern of an automotive head lamp on a road surface is presented. The most critical area where the beam pattern uniformity can be identified is the high intensity region. In this area two evaluating parameters, correlation set, and contrast level are discussed. In a low intensity region of the illumination pattern, the same principle can be applied.

Biomedical instrument systems often must determine whether a container or channel contains liquid or air. Many solutions to this problem have been proposed. Detectors that measure changes in capacitance, conductivity, pressure, ultrasonic coupling efficiency, and optical properties such as absorbance and index of refraction are routinely used. The ideal liquid/air interface detector would be small, inexpensive, reliable, immune to external influences, and chemically and physically compatible with the fluids being interrogated. We present a small liquid detector that uses failure of total internal reflection to indicate the transitions between air and liquid in a fluid conduit or container. The detector is manufactured with the same materials as the channels and chambers themselves, and as such is completely inert to the assay fluids it contacts. It is also relatively immune to external interference such as electromagnetic fields, ambient light, and physical variations in the fluids. It is tolerant of manufacturing variations, and would be inexpensive if manufactured even in moderate quantities. A few critical performance variables are analyzed and results from early testing of the sensor are discussed.

Specific examples of the manufacturing and testing of 1 inch diameter fused silica etalons to fractional arc second tolerances will be discussed. A step by step description of both planetary and spindle polishing processes along with the results of each will be presented. Various testing methods will be described along with the benefits of each.

An alternative technique for blazing sinusoidal-profile gratings is discussed. Whereas conventional blazing techniques use an energetic ion beam to mill through a sinusoidal photoresist profile into a substrate, this technque creates the blazed profile in the photoresist itself, without milling into a substrate. Advantages to this technique are discussed, and grating efficiency curves and groove profiles are shown, before and after ion milling. Simulation results that predict the evolution of the groove profile under ion bombardment are also discussed.

We describe a new fabrication technology for silica-based integrated optical components that employs electron irradiation to modify the refractive index of an amorphous SiO₂ substrate. An asymmetric planar waveguide has been fabricated with this method, with a guide depth of approximately 5 micrometers and a core-cladding index difference on the order of 10^-3. The guide is single-moded over the entire visible spectrum, exhibiting losses of less than 12 dB/cm.

Many lasers and telecommunication devices demand sophisticated optics requiring precise control of the thin film coating process. The advantages of ion beam sputtering in providing a reproducible and accurately controlled process will be discussed, and the use of statistical process control for process refinement will be described. Applications to difficult coatings such as dichroic and trichroic filters for diode pumped solid state lasers, Brewster angle plate polarizers, and edge filters for telecommunication devices will be presented.

This paper addresses the materials, processes, and equipment required for large area micro pattern generation as used in the fabrication of display devices. We will point out some of the processing problems in obtaining zero defect patterns and the types of equipment available to obtain the desired result. We will define large area micro images as patterns generated on substrates larger than 8 by 8 inches, having geometries smaller than 0.008 inches.

Differences in materials and manufacturing processes raise many unfamiliar practical issues for engineering teams changing from low volume glass lens products to high volume plastic lenses. These issues may impact a product from concept to final assembly. A number of such issues are discussed in terms of the authors' experiences.

Today's customer is concerned with cost reduction programs and many "grade" their suppliers by their contribution the these programs. Actually, the time to cost reduce is in the design phase, not after production has begun.

A general description of optical coatings and optical coating processes for glass and plastic is given. This paper details the basic operations and concepts in these areas. A description of coating methods, properties, and possible uses is also given.

Automation and polymer science represent fundamental new technologies which can be directed toward realizing the goal of establishing a domestic, world-class, commercial optics business. Use of innovative optical designs using precision polymer optics will enable the US to play a vital role in the next generation of commercial optical products. The increased cost savings inherent in the utilization of optical-grade polymers outweighs almost every advantage of using glass for high volume situations. Optical designers must gain experience with combined refractive/diffractive designs and broaden their knowledge base regarding polymer technology beyond a cursory intellectual exercise. Implementation of a fully automated assembly system, combined with utilization of polymer optics, constitutes the type of integrated manufacturing process which will enable the US to successfully compete with the low-cost labor employed in the Far East, as well as to produce an equivalent product.

We present a simple definition of similarity between two lamp beam patterns. We present ideas for achieving similarity between a computer simulation of the photometric performance of an automotive head lamp utilizing a free-form reflector (FFR) and the measured hardware photometric performance of the FFR. To achieve similarity, we perturb the computer generated FFR surfaces, which simulates the tolerances generated on the FFR surface during hardware creation. We present computer simulations of the photometric performance of a designed FFR, computer simulations of the photometric performance of the same designed FFR with the 'perturbation', and measure photometric results of the hardware build of the FFR. We utilize our similarity definition to judge the resemblence between the hardware and software results. Finally we present ideas for further uses of the idea of similarity and ideas to further improve the similarity between computer simulations and actual hardware.

A precisely molded acrylic lightpipe is combined with micro-structured films, reflective films, adhesives, and other mechanical components in the construction of a thin, high performance LCD backlight. We focus here on the manufacturing issues associated with the injection molded lightpipe.

A new and improved railroad crossing roundel lens with 12 inch diameter, 0.21 inch thickness using polycarbonate was designed and manufactured. This design meets or exceeds all the recommendations by the Association of American Railroads. The combination of the new design, precision tooling fabrication, and computer-controlled processing technology has allowed this new design to out-perform the existing products on the market.

Today's optical designers face new corporate cultures whose priorities include product performance as only one criteria for success. Designers must also address cost constraints, new and unfamiliar skill requirements, overhead containment, maintenance of profit margins, and staff reductions. Old skills must be applied in new ways. New diamond-turning machine technology has made it possible to construct injection molding tools which combine refraction and diffraction into a single lens element. New polymer materials render the designs to be technically and commercially feasible. The significance of combining refraction and diffraction in a single lens element should not be underestimated, as it will expand the capability of polymer optics beyond its refractive limitations. Use of this technology can restructure domestic optical manufacturing.

The finite element method is used to model the near field of optical components with dimensions on the order of a wavelength. Because open region problems are considered, a conformable local radiation boundary condition is used to truncate the domain and preserve the sparsity of the resulting matrix equation. Diffractive surface examples are presented.

Design issues related to implementation of the scattering optimization method for aperiodic gratings are discussed. Critical design parameters are highlighted and it is shown how their selection affects the final grating structure. The aperiodic grating design technique is then implemented to develop a grating for TE₁₁ to TE₁₁ mode conversion in a circular waveguide at 20 GHz. The length of the grating is 19cm and it has a conversion efficiency of 98.34%.

Diffractive optics are employed as cavity mirrors and intracavity elements in advanced laser resonator designs. The diffractive elements are used to produce custom fundamental mode shapes, discriminate against higher-order modes, and correct thermal aberrations in laser crystals.

A concept for performing spectrum analysis of micro-millimeter wave signals using a novel optical grating approach is described. In the approach an optical wave is temporally modulated with the micro-millimeter wave signal, diffracted from a slanted holographic grating, and focused using a simple lens. The slanted grating acts to convert the temporal modulation into spatial form and the lens produces the power spectrum of the signal in its focal plane. Performance levels projected with a 30 cm grating include resolution of 1 GHz and bandwidth of 200 GHz. Use of in-fiber gratings in a raster format will enable resolution of 2.5 MHz.

We introduce a low cost apparatus utilizing a PIN photodiode receiver and a LED transmitter for frequency domain optical diffusion imaging. We present sample data to demonstrate the system performance and discuss the performance of LED sources and solid state detectors.

We report the observation of intensity fluctuations in a semilinear photorefractive oscillator, consisting of a BaTiO₃ crystal as the gain medium, and a multimode helium-neon laser as its pump source. Depending on the beam geometry and intensity, the periodical fluctuations range on a timescale from a few seconds to several minutes. The behavior is suspected to be caused by a competition of gratings due to the external cavity, crystal-ordinary mirror assembly, the self-pumped phase conjugate mirror due to total internal reflection.

We experimentally demonstrate a method of high-speed optical amplification using degenerate four-wave mixing in a bulk semiconductor CdTe. In our experiment, both transmitted probe and phase-conjugate beams are amplified in accordance with theory. We propose that this approach, which performs optical splitting with gain, can be used with multiple input signals, leading to potential applications including image amplification, information distribution, and optical computing.

In connection with recent theoretical predictions, enhancement of optical second harmonic generation (SHG) by diffractive coupling to the silver surface-plasmon polariton (SPP) mode is shown experimentally to be maximized on a biperiodic corrugated surface. Optimized first- order diffractive coupling of incident radiation to the SPP maximizes resonance enhancement of the surface-localized electromagnetic field. Through the nonlinear susceptibility of the silver surface, an SPP-enhanced, evanescent, second harmonic wave is coherently generated which is selectively scattered into the second harmonic specular order by the second spatial harmonic in the surface profile. Biperiodic surfaces with appropriately optimized spatial harmonic composition are shown to provide enhancements in second harmonic relection of up to 10⁴ over the corresponding flat surface response. Using the hologrpahic technique of Breidne et al. (Fourier blaze holography), biperiodic surfaces were fabricated which consisted of a superposition of an 833 nm fundamental and a phase- and amplitude-controlled second spatial harmonic. The results of angle-resolved SHG experiments are presented along with atomic force microscopy line scans of the surface profiles. The effect of coupling to the SPP mode at both the incident and second harmonic frequencies is also discussed.

Two different demodulation schemes, heterodyne and passive homodyne, were used to determine the frequency response of a polarization-maintaining (PM) optical fiber Mach- Zehnder interferometer for simultaneous measurement of axial strain and temperature. A symmetric 3 by 3 coupler was used to construct the passive homodyne configuration. The heterodyne setup consisted of two acousto-optically modulated lightwaves combined in a 2 by 1 PM fiber coupler. The intensity output of the coupler was demodulated using a digital lock- in amplifier. For a sensing fiber length of 0.54 meters, the frequency response to an axial load was flat up to 2 kHz. A simple rigid rod model of the fiber was developed to estimate the dynamic response.

A localized fiber optic interferometric sensor was embedded in an epoxy plate for the purpose of detecting ultrasound. The sensor embedded was a low finesse Fabry-Perot, which was actively stabilized by tuning the laser frequency. The embedded sensor was used to detect ultrasound generated by a range of techniques which included: i) a 5 MHz piezoelectric transducer, ii) surface laser generated ultrasound, iii) laser generated ultrasound using an embedded fiber generator, and iv) a simulated acoustic emission event (lead pencil break). The frequency content of these signals ranged from the hundreds of kilohertz to beyond 5 MHz. The embedded FOFP was able to detect all of the signals generated, indicating its utility as a broadband ultrasound detector.

Many mines rely on toxic gas sensors to help maintain a safe and healthy work environment. This report describes a prototype monitoring system developed by the US Bureau of Mines that uses light to power and communicate with several remote electrochemical toxic gas sensors. The design is based on state-of-the-art optical-to-electrical power converters, solid- state diode lasers, and fiber optics. This design overcomes several problems associated with conventional wire-based systems by providing complete electrical isolation between the remote sensors and the central monitor. The prototype accurately monitored three remote gas sensors during a two-week field trial in the USBM Pittsburgh Research Center Safety Research Coal Mine.

An experimental study is conducted to evaluate the performance of surface-mounted fiber optic strain sensors for delamination damage detection in eight-layer symmetric cross-ply Graphite/Epoxy composite plates. Teflon film and waxed paper inserts were embedded to model delamination experimentally. Extrinsic Fabry-Perot fiber optic sensors were used to detect such pre-existing delaminations. A fringe-order identification computer program for interferometric optical fiber strain sensors is used for automated strain detection. The responses of surface mounted Fabry-Perot optical fiber sensors due to real-time low velocity impact of rigid spheres with the delaminated and healthy composites were analyzed with respect to the first impact strain peak. The data was validated by measurements of out-of-plane acceleration using conventional piezo accelerometers.

We report two fiber components with the geometrical features etched directly on the fiber cladding. When fibers with etched corrugation structures are stretched or compressed longitudinally, lateral bending occurs. For multimode fibers, the lateral bending leads to attenuation. For two-mode or few-mode fibers, lateral bending causes modal coupling. Taking advantage of these effects, multimode fiber strain sensing elements and two-mode modal couplers are realized. The theoretical models and experimental results of these etched fiber components are presented. For quantitative characterization of the modal couplers, a novel polarization independent LP₁₁ mode stripper was designed and implemented with etched fibers.

Stabilized and near shot noise limited operation of the all-fiber in-line Sagnac interferometer current sensor is demonstrated. We employed a 1.3 micrometers LED as the light source and used a depolarized scheme to suppress polarization and modulator errors.

A novel, working, fiber optic sensor is described which samples scattered light from a flowing white powder stream and monitors the yellowness with very high sensitivity and precision. The monitor does not make a primary measurement of color but is correlated to spectroscopic reflectance measurements of pressed pellets of the powder. It was specificaly developed for, and is used with, purified terephthalic acid, a monomer of polyester. It is, however, clearly applicable to any light scattering power or slurry.

A He-Ne laser beam is guided through an optical fiber, with an audio modulating electric field directly applied to a small part of a relatively long optical fiber. The photodetected lightwaves are modulated by the modulating electric field. The effect of varying the length of the interaction region between the lightwaves inside the fiber and the applied electric field on the observed photodetected signal is investigated. The effect of varying the modulating electric field strength and frequency on the observed nonlinear interaction between the lightwaves and the applied modulating field is also investigated. Results and discussion of the observed effects are presented.

The authors M. Taalbi and T. K. Ishii have found that when a He-Ne laser light is fed to optically transparent glass rods and optical fibers, where an acoustic wave is applied normal to the light beam, the lightwave is modulated by the acoustic wave. The acoustic signal is generated by a piezoelectric transducer, operating in the audio range, with power levels about 1 mW. The detected light contains the fundamental frequency of the modulation signal, with various degrees of phase-shift, frequency doubling and tripling, depending on the location of observation. The lightwave power transmission coefficients are found maximum near the cross-polarization condition, although the light transmission through the samples become minimum at or near the cross-polarization condition. Also, these coefficients are found to be modulation frequency dependent, with values exceeding one, at near cross-polarization. The authors assume that pseudo-amplification has occurred at the previously stated conditions. Theoretical explanation, and models of the observed phenomena have been developed by the authors.

We present a theoretical investigation of the dispersive properties of linearly-chirped fiber Bragg gratings, establishing optimal parameters for the refractive index variation along the fiber length. Employing numerical methods, we show that a 3-cm-long chirped Bragg filter is able to compensate for the dispersion from 100 km of fiber at 1.2 micrometers , recompressing a 20 os Gaussian input pulse from its dispersion-broadened 74.4 ps width back to 20.6 ps. We also prove that a uniform grating profile provides the optimum compression for this input pulse type, confirming recently published claims.

We demonstrate a single-sideband in-fiber frequency shifter using acousto-optic polarization mode coupling. Using 200 mW power to a bulk shear-wave acoustic transducer bonded to a polarization-preserving D-fiber, we have observed power coupling from an optical carrier to an upper sideband, shifted by the 17.5 MHz acoustic frequency. The suppression of the lower sideband, relative to the upper sideband level is at least 10 dB, with the actual sideband level below the noise level of the experiment. This frequency shifting property may be used for the development of in-fiber Bragg cells with improvement of the weak 0.16% observed power coupling efficiency in this work.

Commercially available GaN blue LEDs have been characterized for use as light sources for chemical sensors. These new LEDs are a double heterojunction structure of InGaN/AlGaN that have a peak output at 450 nm. Other groups have investigated these devices for full color displays. This investigation addresses parameters critical to chemical sensors. Several different paramters were characterized including spectra verses drive current, spectra before and after aging, output power verses drive current, and lifetime. The results of this characterization indicate that these devices perform well for some chemical sensors.

Results of 2D numerical simulation of a four-terminal light emitting thyristor in the regime of incomplete turn-off are presented. This regime is characterized by the negative gate current, which is insufficient to turn the device off. A part of the middle p-n junction is reverse biased and blocks the current, whereas the remaining part of the structure is highly conducting and light-emitting. The size of the light-emitting area and the light intensity in this region can be controlled using small gate signals. Gates make it possible also to control the position of the light-emitting region. The utilization of incomplete turn-off principle can be used for light intensity modulation and switching purposes.

In this work, we present a complete numerical simulation of a light emitting diode based on the alloys AlGaN and InGaN. Using material parameters obtained from the literature, a good comparison between numerical simulations and experiments can be achieved. For high brightness in these devices, we would like to have most of the recombination taking place at the dopants in the active region of the device. However, if this region is not doped high enough, we observe that the output power will be saturated. This therefore also gives us a method for finding the concentration of dopants in the active region.

A micro-Raman apparatus was used to detect an object embedded within a scattering medium. The Raman vibrational frequency of diamond in an intralipid scattering medium was detected at different radial distances from the diamond. Scanned images of a single diamond, two diamonds, and a diamond through an aperture are presented. This experiment shows that Raman spectroscopy can be a useful tool in locating and characterizing heterogeneities contained within a scattering medium.

Transmission electron microscope (TEM) techniques are useful for studying the microstructure and chemical composition on nanometer sized areas. Microdiffraction, high resolution TEM imaging and electron energy loss spectroscopy (EELS) are used to investigate the nucleation and growth of diamond films from fullerene precursors. Fine grained diamond films have been deposited on scratched silicon substrates using a microwave plasma consisitng of argon, 1- 10% hydrogen and a fullerene (C₆₀) carbon precursor. These films are exceptionally smooth as measured with an atomic force microscope, with an average surface roughness of 40-100 nm. The growth rate is comparable to that of conventional, hydrogen/methane (CH₄) microwave plasma CVD growth method, which typically results in films with >> 1 micrometers surface roughness. Grain sizes were measured from plain view bright field images from both thin and thick regions of the sample of a film grown from C₆₀ precursors. The smallest and largest grains were 3 nm and 103 nm, respectively; and the median grain size was 13 nm. Cross section TEM images reveal equiaxed diamond grains and continuous nucleation of new diamond grains throughout the growth of the film. Figures show a cross section view, high resolution TEM lattice image of a film grown from C₆₀ precursors. The silicon surface is noticeably rough from the scratching pretreatment. Microdiffraction shows the layer visible between the silicon substrate and the diamond film to be amorphous. The EELS spectrum from this 16 nm thick layer corresponds to amorphous carbon. While EELS spectrum collected from the diamond film corresponds to that of bulk diamond. Argon plasma-assisted CVD diamond films grown from C₆₀ typically nucleate on an amorphous carbon layer on the Si substrate and on diamond residue left from scratching pretreatment. Using fullerene for the precursor, there is continuous nucleation of equiaxed diamond particles throughout the thickness of the film and no columnar grain growth is observed.

Recent research by Doane, Yang, and Chien demonstrated the use of cholesteric liquid crystals in multiplexed, high resolution, reflective diplays. These materials utilize the bistability of the cholesteric planar and focal conic states for displays with a colored image on a black background. Many commercial applications of these materials, such as electronic books and newspapers, portable faxes and personal data assistants, require, or at least prefer, black-on- white images. We report on relatively high polymer content (equalsV 20% by weight) dispersions of cholesteric liquid crystals that produce a white, reflecting, planar state. The polymer network appears to form cholesteric domains with varying pitch lengths resulting in planar states that reflect in the red, green, and blue portions of the spectrum. Utilizing a black absorbing layer behind a display using these materials offers white images on a black background, or vice-versa.

We report on the optical reflective properties of the planar texture of cholesteric liquid crystal displays. The cholesteric liquid crystal is made bistable by either dispersing a low concentration of polymer or by treating the cell substrate surfaces. We determine the role that the polymer network and surface treatment has on the reflective properties as a funciton of viewing angle using both collimated and diffuse illumination. Both the polymer network and surface treatment have the effect of distributing the orientation of the cholesteric helix axes about the cell normal. Theoretically we characterize these cells by this distribution.

The availability of LCD panels such as found in laptop computers has permitted the establishment of new devices that permit display of computer monitor and video images. These LCDs were initially employed in panels that were placed on the stage of overhead projectors. They have now been integrated into complete projection products and are beginning to replace the 3-CRT projectors. The current technology of the LCD components will be discussed along with their unique properties when incorporated into a projection device. LCD technology has been evolving at a phenomenal rate. A full color LCD of the type used for a computer display became available only four years ago. Today, LCDs of exceptional quality and resolution exist. Still there are special operational properties, such as display control, transmission, contrast, color level, and response time, which must be addressed. Furthermore, there are both single large LCDs and the three monochrome color elements which when combined with dichroic optics yields a full color projector. Incorporating such LCDs with high output lamps results in reasonably bright images. These projection devices are replacing the film based presentation technology and providing the tools for computer generated, electronic multimedia presentations.

A new technique for real-time optical character recognition using a joint transform correlator is proposed. This technique employs a special feature-extracted reference image for detecting a wide range of characters in one step. The proposed system also shows feasibility for feature- extracted pattern recognition.

A display for presenting real-time 3D video or computer generated images is described. The display consists of an LCD panel, two HOEs and a white light source.

Color holograms recorded in panchromatic, single-layer, ultra-high-resolution, silver-halide emulsions have been previously reported. Color reflection holography presents no problems with regards to the geometry of the recording setup, but the final result is highly dependent on the recording material used. The processing of such emulsions is critical in order to obtain high diffraction efficiency and good color rendering. In particular, preventing emulsion shrinkage is extremely important. The recording procedure and the processing steps will be described.

Holography with very short pulses, e.g., in the femtosecond regime, has some significant capabilities. Also, it has some special problems. These capabilities and problems are described, and some solutions are offered for the problems.

The practical application of electronic holography requires the use of fiber optics. The need of employing coherent fiber optics imposes restrictions in the efficient use of laser light. This paper proposes a new solution to this problem. The proposed method increases the efficiency in the use of the laser light and simplifies the interface between the laser source and the fiber optics. This paper will present the theory behind the proposed method. A discussion of the effect of the different parameters that influence the formation of interference fringes is presented. Limitations and results that can be achieved are given. An example of application is presented.

This paper presents a simple method to design a recursive filter for processing ESPI fringe pattern to obtain a quality results according to the frequency of the fringe. This method of speckle noise reduction has the advantages of both less computation time and improved results.

This paper employs the iteration algorithm proposed in reference to process shearography fringe patterns. The algorithm processes speckled fringe pattern several times until the quality of the processed fringe pattern is acceptable. Examples and description of the algorithm are presented.

In this paper we propose a new method of shearing in an electronic speckle pattern shearing interferometry using holographic gratings. The system we are proposing consists of two parts: 1) to image the object onto an intermediate ground glass, and 2) the image on the ground glass is in turn imaged onto the photosensor of a CCD camera. A holographic grating placed in front of the ground glass screen is used for shearing the two images and for introducing phase stepping.

In this paper further discussions on some new methods proposed in an earlier paper on phase stepping interferometry is detailed. These studies reveal that certain algorithms are more sensitive to miscalibration and nonlinearity and can be better candidates to detect the presence of phase shifter miscalibration or nonlinearity. A simplified approach to the understanding of the error and its reduction/elimination in phase shifting interferometry are also presented.

A shear interferometer using wave mixing and angular multiplexing techniques in a BaTiO₃ crystal is investigated. By varying the angle between the pump beam and the probe beam, two separate holograms in the form of two index gratings are formed in the crystal. Each grating has a different magnitude and direction in the k-space. Reconstruction is attempted with both the same wavelength as that of the writing beams and with a slightly different wavelength. During the reconstruction process the crystal orientation is adjusted such that the phase-matching condition is satisfied simultaneously for both gratings. As a result of the reconstruction two sheared wavefronts are obtained. A theoretical description, results of theoretical investigations and an experimental configuration of this interferometer are presented.

Real-time interferometry of vibrating specimens is achieved using two-wave mixing in Bi₁₂SiO₂₀ (BSO). By using anisotropic self-diffraction in the photorefractive crystal, despite the small diffraction efficiency of the photorefractive grating, good fringe visibility and good singal to noise ratio are achieved by synchronizing the illumination with the CCD camera. In addition, the application of an external electric field is used to enhance the diffraction efficiency of the photorefractive grating. This greater efficiency makes possible the visualization of large diffusely scattering specimens. It is shown that the good SNR of the resulting holographic interferograms enables phase unwrapping which allows for quantitative deformation analysis. The performance of the photorefractive interferoemeter is compared with the performance of an electronic speckle pattern interferometer.

This paper presents a method for measuring time dependent displacement using a dual speckle pattern acquisition and single camera TV holographic system. Two speckle patterns with 90 degree phase difference are acquired simultaneously. As the displacement changes with respect to time, a series of dual speckle patterns can be digitized into a computer image processing system. A Fourier transformation is used to remove the DC terms of the speckle patterns. A phase unwrapping algorithm is employed to calculate the phase distribution related to the time dependent displacement. The theoretical concept together with experimental simulation are presented.

This paper presents a method for measuring the complex mode of a vibrating structure using two separate dual reference stepped strobe phase (0, 90 degrees) holograms and a new spatial phase shifting technique during reconstruction. Spatial phase shifting is accomplished by scanning a pencil of a laser conjugate reconstruction beam in four equal steps across the hologram; four shifted phase real images are obtained for each of the two holograms and each set of four images is processed to get the unwrapped phase distribution. The vibration amplitude and phase are then obtained using these two phase distributions. The new spatial phase shifting technique greatly simplifies data acquisition and optical readout since only a single reconstruction beam is used to reconstruct both wavefronts for each hologram. The theory and experimental results are presented.

The paper presents a novel method for recognizing raised or indented characters or patterns on industrial samples by using a combination of moire interferometry technique with optical character recognition (OCR) and pattern recognition. Patterns recognized with this method are of low contrast, and conventional recognition schemes require complex optics and lighting. Raised characters on tires, vin code tags, credit cards, indented characters on metal, wrinkles on skin, and embossment on buttons are some examples. The proposed method uses the moire interferometry technique to obtain a gray scale image of patterns such that their heights are represented in gray scale. This eliminates the need for special optics for each application. 3D images obtained as above, are processed by three sets of algorithms: 1) analytical geometry, 2) pattern recognition, and 3) character recognition. The analytical geometry algorithms consist of constrained and unconstrained fitting methods for scattered data, and transformations between different spaces. The pattern recognition methods consist of feature extraction based on scatter matrices, and classification based on hierarchic classification methods. The OCR algorithm employs gray scale correlation. Extension experiments are conducted to support the method.

The detection of product discoloration is very important to many industrial applications, such as the inspection of vegetables, fruits, grains, pills, etc. There are currently two methods for detecting the color change. The first method is using an optical sensor to measure the amount of light energy that is reflected by the color surface. The other method determines the change in product color with the black and white image. The performance of these two methods suffers, due to the lack of the color information and poor resolution. A high-speed automated color sorting vision system is designed and developed using the IDAS^TM (imaging development and application system) real-time image processing system to provide true discoloration detection. A color line-scan camera, which detects the amount of light in the red, green, and blue color spectrum, is used to obtain color information from the product surface. The conbinations of information from these three color channels cover almost all of the color variations that human eyes can distinguish. The information used by this vision system is at least three times of that by the black and white system and the defect resolution required can be achieved by selecting the proper optics and the camera sensor size. The use of the pipeline real-time image processing hardware and the implementation of the image processing algorithms enable the system to detect the discoloration at very high speed. The detail descriptions of the configuration of the vision system, optics selection, image processing algorithms, hardware control, and communication will be introduced in this paper.

Reconnaissance from unmanned platforms is currently of interest to DoD and civil sectors concerned with drug trafficking and illegal immigration. Platforms employed vary from motorized aircraft to tethered balloons. One appraoch currently under evaluation deploys a TV camera suspended from a parafoil delivered to the area of interest by a cannon launched projectile. Imagery is then transmitted to a remote monitor for processing and interpretation. This paper presents results of imagery obtained from simulated parafoil flights in which software techniques were developed to process-in image degradation caused by atmospheric obscurants and perturbations in the normal parafoil flight trajectory induced by wind gusts. The approach to capturing continuous motion imagery from captive flight test recordings, the introduction of simulated effects, and the transfer of the processed imagery back to video tape is described.

Electronic imaging technology has advanced to the point where sophisticated wafer-scale imaging devices are on the verge of being fielded for demanding military reconnaissance applications. High-density, 4-Mpixel, and 25-Mpixel silicon CCD detectors have been fabricated and integrated into a proof-of-concept reconnaissance sensor, the CA-260. One unique feature of this sensor is its ability to compensate for image motion during very high- speed imaging conditions. This paper discusses the patented detector architecture and system design considerations of the CA-260, and review the results of flight tests conducted on various military aircraft during 1993 and 1994.

We present a new algorithm for the discrimination of remote objects by their surface structure. Starting from a range-azimuth function (RAF), we formulate a range-azimuth matrix whose largest eigenvalues are used as discriminating features to separate object classes. A simpler, competing algorithm uses the number of sign-changes in the RAF to discriminate between classes. While both algorithms work well on noiseless data, an experiment involving real data shows that the eigenvalue method is far more robust with respect to noise than the sign change method.

We have developed a vision application using the IDAS^TM (imaging development and applciation system) real-time image processing system to inspect discrete stamped metal parts. Specifically, the stamped clutch plates coming through the manufacturing line at speeds higher than 6000 parts per hour are to be inspected for defects on both surfaces of the clutch plate as well as defects on the profile of the clutch plate with dimensional accuracy in mils. The conventional area scan approach will not be practical die to the resolution requirements and the processing speed. We have developed a solution based on a 2048 CCD linescan camera and a conveyor system to transport the plates under the camera. This real-time multitasking system deals with all the complexities one normally encounters during the implementation of a real-time machine vision system. This paper will describe the application and the design strategies that made the solution both practical and economical.

In most artificial neural network applications, (e.g. pattern recognition) if the dimension of the input vectors is much larger than the number of patterns to be recognized, generally, a one- layered, hard-limited perceptron is sufficient to do the recognition job. As long as the training input-output mapping set is numerically given, and as long as this given set satisfies a special linear-independency relation, the connection matrix to meet the supervised learning requirements can be solved by a noniterative, one-step, algebra method. The learning of this noniterative scheme is very fast (close to real-time learning) because the learning is one-step and noniterative. The recognition of the untrained patterns is very robust because a universal geometrical optimization process of selecting the solution can be applied to the learning process. This paper reports the theoretical foundation of this noniterative learning scheme and focuses the result at the optimal robustness analysis. A real-time character recognition scheme is then designed along this line. This character recognition scheme will be used (in a movie presentation) to demonstrate the experimental results of some theoretical parts reported in this paper.

The extraction of facial features is a fundamental and crucial step in most face detection and recognition systems. Here a set of approaches are proposed for extracting internal and external facial features in a grey-level images containing single or multiple faces of various sizes at different locations without any restriction on the background. These approaches are distinctive in several aspects: the use of some rarely exploited photometric properties as the basis for facial features extraction, a novel metrics on radial symmetry of gradient orientations, an inhibitory mechanism for extracting internal facial features, and a simple mechanism for external ones.

The train energy model (TEM), a general train simulator developed at the Association of American Railroads, is widely used in the railroad industry. Recently, a new train controller, the general automatic train-controller (GAT), has been developed for TEM. In the GAT, the 'intelligence' or 'expertise' is a set of 'if-then' train-handling rules in an external file. The expert system for train control presented in this paper is a slightly simplified version of the GAT. The main thesis of this paper is: simple, unchained rules are adequate for complex train control. Thus, an 'inference engine' using forward chaining is not required.

This paper introduces a relatively simple multimode fiber optic sensor built for on-line use and a package that multiplexes two fiber sensors. The multiplexing is achieved by the random nature of the multimode fiber output intensity variation (so called speckle pattern). The demultiplexing is performed by a neural network. The dynamic range of each sensor is 0.8- 120 micrometers over an effective length of 11 m. The sensors operate in 68-94 degrees F.

Today neural networks are being adopted as an alternate method of solving complex pattern recognition/classification problems. Information regarding performance measure is critical in evaluating the capacity of this system in performing recognition/classification tasks. Currently this information is obtained using unstandardized empirical techniques. This study will attempt to devise a methodical procedure to qualitatively predict the performance measure of all neural network recongition classification systems governed by a set of ordinary differential equations. The determination of this characteristic will be made through the use of specific analytic methods in mathematics. Dynamical systems can therefore be qualitatively analyzed, and issues regarding existence of parasitic limit points can be more effectively addressed.

Automatic control of traffic signals in many cities are often based on a constant green-to-red cycle. The time period for green light (or red light) to be on is fixed, and is determined based on a stochastic model. In some situations, a human operator, or a fixed timer, may change the time period dyring time intervals where there is heavy traffic--typically during commuting times. Although in today's traffic controllers, the time period for green light can be changed in certain circumstances, in general they are not able to adjust dynamically to traffic changes. This paper presents the design of a controller that considers the intersection of two two-way streets and is able to adjust to environmental changes. Simulation of the proposed controller has produced results more in line with what we expected. In many cases, the proposed system has produced better results than systems that are based on a constant green-to-red cycle.

A new class of quality measures is emerging in the literature for image compression. The measures in this class are graphical, producing multidimensional output that provides more information than scalar measures like MSE. Eskicioglu charts, a recent addition, computes three features for various frequency ranges, and displays them in a bar chart. Reflecting the human visual response, it is able to specify the amount, type, and distribution of error in compressed images. In this paper, we discuss adding one more dimension to Eskicioglu charts for further improvement. The feature corresponding to the new dimension is the end-of-block disturbances which increases for blockiness and decreases for blurriness.

A dual wavelength, imaging microdensitometer, with specific application to measurements of DNA ploidy and to monoclonal antibody related immunoenzymatic staining procedures, is described. The two wavelengths used were 500 nm and 620 nm, each with a 20 n bandpass to reduce glare and to provide independent sensing of typical immunohistochemical two-color stain spectra. The optical arrangement is such that the original multicolor stained image is separated into two components by two filters and projected simultaneously onto two separate CCD sensors that are in optical registration for the same spatial regions in each image. Independent measurements of individual stained subcomponents of each cell can be made, and the entire image is reconstructed digitally and displayed showing both of the sensed subcompartments in one image. Optical methods of calibration for obtaining accurate densitometry in two colors simultaneously are discussed, as well as various applications related to performing assays in surgical pathology for cancer prognosis.

Until recently, the ability to measure the changing oxygen gradients in perfused tissues in response to metabolic demand, has been limited to point-measurements and/or averaged A-V oxygen differences during perfusion using oxygen electrodes. With the recent introduction of novel phosphorescent probes specifically quenched by oxygen, the ability to spacially map oxygen gradients in real-time may offer new insights into the dynamics of microvascular design and supply. Accordingly, this paper provides initial image data on Langendorff perfused rat hearts wherein the relative change in phosphorescent intensity of Pd-meso-tetra(4- carboxyphenyl)phorphine (2micrometers ) as the reporter probe, is quantitatively related to spacial oxygen gradients as seen on the left-ventricle during changing gassing conditions. Digital image analysis (frame advance), after proper calibration and alignment, provides images which can be usefully interpreted. Clinical applications of such emerging technologies could have wide-spread diagnostic applications not only as applied to the coronary bed, but other tissue surfaces displaying various degrees of aschemia and/or hypoxia.

This paper describes an efficient method of segmenting cell nuclei from complex scenes based upon the use of adaptive region growing in conjuction with nucleus-specific filters. Results of segmenting potentially abnormal (cancer or neoplastic) cell nuclei in Papanicolaou smears from 0.8 square micrometers resolution images are also presented.

Using infrared sensitive cameras and on-line image processing, pupil diameters of awake, eyes-open subjects were measured. Concurrently, electroencephalography (EEG) was monitored and power in the delta (0.5-4Hz), theta (4-8Hz), alpha (8-12Hz), sigma (12-16Hz), beta₁ (16-26Hz), and beta₂ (26-50Hz) bands were calculated. Pupil diameter and EEG power measured were found to be significantly correlated. Other EEG measures including relative beta (defined here as [power of 16-50Hz]/[power of 4-50Hz]) and centroid frequency of the 4-50Hz band were also found to be significantly related to pupil diameter.

The design of a two-channel dynamic dermofluorometer to monitor skin perfusion using a fluorescent dye is reported in this paper. The hardware of this system consists of an optical module (flash lamp, optical fibers, dichroic beamsplitters, and optical wavelength filters), an electrical module, and a Macintosh computer. This system is controlled by software written using LabVIEW and MPW C running with the support of a Lab-NB multifunction board. This instrument is capable of real-time data acquisition, curve fitting, and display. Two fluorescent dyes, carboxyfluorescein and sodium fluorescein, were used in prelimiary animal studies (6 rats). Each dye was injected intravenously and its signal collected by an optical fiber bundle (diameter 3.2 mm) placed on the skin. The washin time constants were similar for both dyes (approximately 5 min.), but sodium fluorescein exhibited a much slower clearance time constant (110 min. vs. 65 min.). The instrument was tested on four normal human volunteers. The washin time constant of sodium fluorescein (1 mg/kg) for two skin sites on the forearm was measured on each subject (average 2-3 min.). These results demonstrate the potential of this instrument for performing dynamic fluorescence measurements. Future studies on diabetic patients are planned at the Danville, Illinois VA Medical Center.

We have developed a new approach to perform body surface Laplacian imaging of cardiac electrical activity by estimating the body surface Laplacian electrocardiograms from potential electrocardiographic signals. This paper describes a numerical algorithm to estimate the Laplacian electrocardiogram from the potential electrocardiograms. Computer simulation studies have been conducted to evaluate the accuracy of the developed approach in aspherical torso volume conductor and in a realistic geometry torso volume conductor. Our preliminary simulation results demonstrate the feasibility of performing body surface Laplacian imaging of cardiac electrical activity from body surface potential electrocardiograms.

Development of the DICOM standard and incremental developments in workstation, network, compression, archiving, and digital x-ray technology have produced cost effective image communication possibilities for selected medical applications. The emerging markets include modality PACS, mini PACS, and teleradiology. Military and VA programs lead the way in the move to adopt PACS technology. Commercial markets for PACS components and PAC systems are at $LR400 million growing to $LR500 million in 1996.

SPECT can potentially be used for quantitative imaging of in vivo 3D radiopharmaceutical distributions. Attempts for accurate quantitation in 3D SPECT images have been compromise not only by the physical effects of photon attenuation, distance-dependent spatial resolution, and scattering, but also by the lack of effective and efficient methods that will correct for these effects. In this work, we introduce a one-step method that can effectively compensate for the effects of photon attenuation and distance-dependent spatial resolution in 3D SPECT. The correction for these effects requires only a very limited amount of computation in addition to that for 3D reconstruction and hence has the potential for routine clinical application. We use both computer-generated simulations and real data to validate the approach. The results demonstrate that the proposed one-step compensation method results in reconstructed 3D SPECT images with good quantitative information.

Forty-six thousnad women die each year in the US from breast cancer. Mammography is the best method of detecting breast cancer and has been shown to reduce breast cancer mortality in randomized controlled studies. Clustered microcalcifications are often the first sign of breast cancer in a mammogram. The use of a second reader may improve the sensitivity of detecting clustered microcalcifications. Our laboratory has developed a computerized scheme for the detection of clustered microcalcifications that is undergoing clinical evalution. This paper concerns the feature analysis stage of the computerized scheme, which is designed to remove false-computer detections. We have examined three methods of feature analysis: rule-based (the method currently used in the clinical system), an artificial neural network (ANN), and a combined method. To compare the three methods, the false-positive (FP) rate at a sensitivity of 85% was measured on two separate databases. The average number of FPs per image were: 0.54 for rule-based, 0.44 for ANN, and 0.31 for the combined method. The combined method had the highest performance and will be incorporated into the clinical system.

Conventional magnetic resonance images are reconstructed by Fourier transformation and have uniform spatial resolution across the entire field of view (FOV). This paper describes a way of creating MR images which have higher spatial resolution in some areas than others. High resolution imaging can be confined to just those areas where it is necessary to resolve strong edges without truncation artifacts. Such locally focused images can be acquired in less scan time than required to image the entire FOV with uniformly high resolution. After the user specifies the spatial resolution in each portion of the FOV, the algorithm automatically generates image basis functions which oscillate most rapidly in the regions with highest resolution. Images are reconstructed by summing image projection onto these basis functions. These projections are calculated from a subset of the usual phase-encoded signals required to create a uniformly well-resolved image. The algorithm also determines which phase-encodings are optimal for this purpose, and these are usually nonuniformly scattered in k-space. Thus, both data acquisition and image reconstruction are optimized. Functional and interventional imaging may benefit from this technique, which makes it possible to acquire a rapid series of dynamical images which have high resolution in areas of expected changes and lower resolution elsewhere. Spectroscopic images may be improved by using high resolution in the neighborhood of sharp edges which might otherwise cause truncation artifacts.

Direct inverse Fourier transform reconstruction methods once had a reputation for generating reconstructions with intolerable artifacts, but since 1980 the methods have been shown to be robust, with artifacts no worse than those in filtered backprojection. This paper shows how such a method was appied to nuclear medicine and tested. The emphasis is on speed of execution, signal-to-noise vs. resolution, and reduction of artifacts below the level of the noise.

In calculating projections for iterative SPECT, a particularly time consuming part of the calculation is convolution of the activity at each depth by the point-spread function, using FFTs. We find that space-based convolutions with a short-range kernel are equivalent and faster, if one uses the central limit theorem to generate a gaussian, depth-dependent point- spread function. This technique is applied to forward and backward projections.

We propose a new algorithm for restoring quantum limited images. The algorithm is based on projection methods but uses constraint sets in which set membership is based on probabilistic measures. Such constraints can be regarded as soft, as opposed to hard constraints in which less latitude is given in defining set membership. We show that the restoration of quantum- limited images has certain similarities to estimating a probability density function. We apply the algorithm to a widely used image phantom and demonstrate that the processed image features less noise without blurred edges.

Positron emission tomography (PET) is a medical imaging modality which produces valuable functional information, but is limited by the poor image quality it provides. Considerable attention has been payed to the problem of reconstructing images in a way that produces better image resolution and noise properties. In dynamic imaging applications PET data are particularly noisy, thus preventing successful recovery of spatial resolution by signal processing applications. In this paper we show that smoothing of image data using a low-order approximation along the time axis can greatly enhance restoration performance.

Videostroboscopy is an examination method during which a video-recording of the vocal folds can be obtained. This examination is very important because it yields a permanent record of the moving vocal folds and it allows the diagnosis of abnormalities which contribute to voice disorders. In this paper a new algorithm based on simulated annealing (SA) is used to register/match the videostroboscopic images of the larynx. This algorithm operates on the contours representing the vocal folds. The matching process is done through minimizing a cost function, which consists of two parts. The first captures the requirement that the distance between two pixels that are matched should be minimized, while the second part requires the smoothness of the displacement vector field. Four parameters are used to characterize the SA algorithm: 1) the acceptance ratio (chi) , which is set to give a high initial temperature (Tau) ₀ to start with, 2) a small positive number (delta) which controls the decrement of the temperature, 3) the length L of the Markov chains that brings the system to equilibrium at each temperature, and 4) a stopping parameter (epsilon) S that defines the final state/configuration of the system. The performance of this matching algorithm is demonstrated on simulated and real videostroboscopic images. It is shown that this algorithm is successful in matching videostroboscopic images when the deformation is severe.

In this paper, we present a model that is used to improve the resolution of autoradiographic images. The model involves a point spread function (PSF) due to the radiated pattern of emitted photons combined with a signal-dependent noise source due to the granularity of x-ray recording film. A theoretical expression for the PSF is presented, and experimental measurements are performed using ⁵¹Cr microspheres. An iterative regularized image restoration algorithm is developed using a weighting matrix to incorporate the signal- dependent nature of the noise. Since information about the original undegraded image is not completely available, we make use of a regualtization functional that is updated at each iteration to optimize the solution process. Our experimental results indicate that the resolution of autoradiographic images is improved by 43% using this algorithm.

This paper presents an interpolation method that uses the contours of organs as the control parameters to recover the intensity information in the physical gaps of serial cross-sectional medical images. In this method, active contour models are used for generating the control lines required for the field morphing algorithm, which were previously manually specified. Contour information derived from this segmentation pre-process is then further processed and used as control parameters to warp the corresponding regions in both input data slices into compatible shapes. In this way, the reliability of correspondence to locate different segments of the same organs is improved and the intensity information for the interpolated intermediate slices can be derived more faithfully, To reduce the high time complexity for calculating the image warp in the field morphing process, a hierarchical decomposition process is proposed. In comparison with the existing intensity interpolation algorithms, which only consider corresponding points in a small physical neighborhood, this method warps the data images into similar shapes according to contour information to provide more meaningful correspondence relationships. The results show that this method generates more realistic and less blurred interpolated images especially when the lcoal intensity variation is significant.

The Bayesian approach that employs the concepts of cliques and line sites in its Gibbs prior provides the potential of realistic and objective characterization of boundaries between different regions exhibiting intensity variations in the reconstructed images. In this work, we develop an improved Bayesian approach for accurate detection of boundaries by introducing symmetric cliques and new types of line sites as well as a new calculation scheme. This improved Bayesian approach has been applied to positron emission tomography data from both computer simulations and patient studies. The results demonstrate that the new cliques, line sites, and calculation scheme can enhance the boundary detectability of the Bayesian approach and yield realistic boundaries and hence improved reconstructed images.

Object recognition is one of the prime probems of computer vision. One way of extracting information is to compute the Gaussian curvature for the given surfaces. The algorithm uses discrete approximation using triangularization methods to compute Gaussian curvature. The images are initially broken down into different segments and the Gaussian curvature for each pixel in the segment is computed with respect to its eight neighboring pixels. These computed values are then converted into intensity format for graphical visualization. The images with improved edge information have been taken from previous work. Synthetic images containing signal object scenes have been tested.

This paper describes an image segmentation technique in which an arbitrarily shaped contour is deformed stochastically until it fits around an object of interest. The evolution of the contour is controlled by a simulated annealing process which causes the contour to settle into the global minimum of an image-derived 'energy' function which is designed to be small when the contour is near the border of objects similar to the target. The nonparametric energy function is derived from the statistical properties of similar previously segmented images, thereby incorporating prior experience. Since the method is based on a state space search for the contour with the best global properties, it is stable in the presence of image errors which confound segmentation techniques based on local criteria such as connectivity. However, unlike 'snakes' and other active contour approaches, the new method can handle arbitrarily irregular contours in which each inter-pixel crack represents an independent degree of freedom. The method is illustrated by using it to find the brain surface in magnetic resonance head images, to identify the epicardial surface in magnetic resonance cardiac images, and to track blood vessels in angiograms.

Diagnostic medical imaging often contains variations of patient anatomies, camera mispositioning, or other imperfect imaging condiitons. These variations contribute to uncertainty about shapes and boundaries of objects in images. As the results sometimes image features, such as traditional edges, may not be identified reliably and completely. We describe a knowledge based system that is able to reason about such uncertainties and use partial and locally ambiguous information to infer about shapes and lcoation of objects in an image. The system uses directional topographic features (DTFS), such as ridges and valleys, labeled from the underlying intensity surface to correlate to the intrinsic anatomical information. By using domain specific knowledge, the reasoning system can deduce significant anatomical landmarks based upon these DTFS, and can cope with uncertainties and fill in missing information. A succession of levels of representation for visual information and an active process of uncertain reasoning about this visual information are employed to realiably achieve the goal of image analysis. These landmarks can then be used in localization of anatomy of interest, image registration, or other clinical processing. The successful application of this system to a large set of planar cardiac images of nuclear medicine studies has demonstrated its efficiency and accuracy.

Information theory indicates that coding efficiency can be improved by utilizing high-order coding (HOEC). However, serious implementation difficulties limit the practical value of HOEC for grayscale image compression. In this paper we present a new approach, called binary-decomposed high-order entropy coding, that signifucantly reduces the complexity of the implementation and increases the accuracy in estimating the statistical model. In this appraoch a grayscale image is first decomposed into a group of binary sub-images. When HOEC is applied to these sub-images instead of the original image, the subsequent coding is made simpler and more accurate statistically. We apply this coding technique in lossless compression of medical images and imaging data, and demonstrate that the performance advantage of this approach is significant.

A progressive contour model is developed based on the idea of deforming the contour from an initial shape as a source of prior knowledge by minimizing a defined contour energy to extract a desired contour from images. This model differs from active contour models (or snakes) in that the internal component of the contour energy is used to impose the smoothness constraints not on the shape of the contour but on the displacements of deformation, and the external component of the contour energy is used to locate the correspondence for the contour through a specified local correspondence mapping. A sequence of deformations is determined by repeatedly deforming and updating the initial contour. It is shown that the contour deformed by this sequence will smoothly and progressively approach a well-defined contour. Finite- element methods, multigrid algorithms, and unconstrained optimization methods are employed to implement this model. This approach offers several attractive advantages including a good convergence rate, the adaptation of the smoothness constraints and the adoption of a globally convergent algorithm. Experiments are conducted on real images to evaluate the performance of a progressive contour program, and a computational complexity in the order of O (lnN) is verified.

This paper is aimed at improving, through artifact rejection, the performance of neural network evoked potential (EP) classifiers designed to detect match/mismatch conditions. A cluster analysis approach is formulated to identify artifacts that occur in the signals used for training the neural network classifiers. The clustering based artifact detection algorithm uses a distance measure resulting from a nonlinear alignment procedure designed to optimally align EP signals. Match and mismatch EPs collected for network training are clustered and the identified artifact signals are excluded from the training set. Artifacts that occur during testing are also identified and rejected by including an additional output in the neural net classifier for the artifact class. Preliminary experiments conducted show significant improvements in classification accuracy when the proposed artifact rejection methods are incorporated in the training and testing phases of a neural network EP classifier.

This paper describes the application of high-resolution, color imaging acquisition, and processing to the inspection of industrial coatings on bridges and other structural elements. Timely inspection of industrial coatings has been identified by state Department's of Transportation and coating inspectors as critical to long-term, cost-effective maintenance protective coatings.

In any electron optical system, compensating for the lens aberrations is the key to improving the spatial resolution. Ofall the dominant aberrations, only the spherical aberration and chromatic aberration continue to limit the performance ofmicroscopes. In the plane ofthe Gaussian image, these aberrations are given by: spherical aberration: ö = Cs a chromatic aberration: ö C a öV/V, Where C5 and C are the coefficients of spherical aberration and chromatic aberration, respectively. a is the divergent angle ofthe electron path with respect to the axis. 8V is the energy spread ofthe electron source and V is the electron energy, in eV. Several systems, such as a magnetic sextupole and quadrupole configuration [1,2], have been proposed to correct these aberrations. Recently, Crewe proposed a simple "mirror corrector" system for correcting them by reversing their coefficient signs [3]. However, it will be some time before any ofthese proposed systems will prove to be applicable on a practical basis. In the meantime, it may be profitable to concentrate on how to choose the best operating parameters for any existing system, in order to optimize the resolution. The resolution of a system is calculated by combining the effect ofthe dominant aberrations with the effects of diffraction and defocus: diffraction: 8d 0.6 lAJa, defocus: C5 a2c where X =(15O/V)' A isthe electron wavelength, and c is the defocus parameter, defined in this paper as 0 at the Gaussian image plane and 1 at the marginal plane. The optimum resolution is achieved by choosing a and to attain the minimum combined aberration. Conventionally, a is chosen to be the largest possible divergent angle, which is the angle determined by the aperture size. This considers only the worstcase situation and ignores the fact that the path of the optical ray is critically dependent on a. Simply because there are some large divergent angles does not mean necessarily that those contributions dominate the formation of the electron probe. Ignoring the possible range of a results in the loss of all the details of the probe pattern. This prevents one from deriving any information about the resolution from the internal details of the probe. Calculations based on a single, maximum value of a can be expected to be conservative and to underestimate the power of the microscope. Recently, Rempfer and Mauck investigated the interior intensity pattern of an electron probe by ray tracing the path ofeach individual electron[4J. Their results showed that the intensity distribution ofthe probe varied greatly in different defocus planes, and was extremely nonuniform in every plane. This paper presents a study in which a numerical simulation is used to determine the probe pattern resulting from the spherical aberration, chromatic aberration, source size, and defocus. The probe pattern in different defocus planes is then convoluted with two types of simulated samples in order to directly investigate the possible resolution. The results show that the conventional calculations ofresolution may not be appropriate for all circumstances. Also, the optimal operational parameter can be very different, depending on what information the operator wishes to get from the microscope.

Holographic interferometry is a well established technique for accurate measurements of mechanical displacements. In cases where no prior information is available on the lcoations of zero displacement, there has not been a method which allows one to determine the location of the zeroth order interferometric fringe. We wish to demonstrate that by using three wavelengths to record color interferometric holograms, the zeroth order fringes are easily identified.

Transcranial cerebral oximetry (TCCO) is a techniqe that evaluates saturation of oxygen in the underlying area of the brain by the noninvasive method of near-infrared spectroscpoy. Human tissues are generally transparent to light in the near-infrared range, so the light of this range (650 nm-1 100 nm) easily penetrates tissue to a depth of several centimeters. The light is partially absorbed by natural chromophores. In the human brain the predominant chromophores are oxygenated (HbO2) and deoxygenated (Hb) hemoglobin. The difference in the absorption spectra between the HbO2 and Hb yields the ratio of the oxygenated to total hemoglobin (the oxygen saturation) in the area of interest. The algorithm used to calculate the oxygen saturation of brain tissue using a multiple-detector system was previously described and validated in a number of animal and human studies. The use of cerebral oximetry in neurosurgery has been described. We discuss some observations from our experience with TCCO.