The use of absorbing baffles for blocking scattered light in space-borne telescopes raises a major problem. The absorbed stray light becomes heat, which in the space environment cannot be dissipated by air convection. This is particularly important for those telescopes operating in the infrared, because the warm baffles will reradiate at those wavelengths. The conventional solution to this problem is the use of cryogenic cooling. We present here an alternative system of reflective baffles that reflect back out of the entrance aperture a preponderance of the stray light. Some criteria are also presented that will aid in the location and the distribution of the individual vanes.

The detector in a background-limited infrared telescope receives less power due to the thermal emission from telescope components than from the background. The Arizona Program for the Analysis of Radiation Transfer has been used to set the temperature of the telescope components so that each component contributes less than 10^-13 W of thermal power to the detector in the Space Infrared Telescope Facility long-wavelength configuration. The secondary mirror assembly, seen from the detector, is the most critical emitter even at 3 K.

Methods for grinding and polishing IR transmitting materials are described and contrasted with better known methods for making glass optics. Because many IR optical systems use aspheric surfaces, single-point diamond turning (SPDT) is suggested as the most prudent method of making both reflective and refractive aspheric surfaces. For cases in which sufficiently good surface finish cannot be achieved with SPDT, either for system use or for testing with visible-light test equipment, a postpolishing technique that uses fine diamond paste and synthetic polishing pads is given as an alternative method. Finally, some practical aspects of good mechanical and optical engineering relating to SPDT are reviewed. By taking a few simple precautions and by building in appropriate optical alignment and mechanical mating surfaces, a difficult test and assembly situation can be reduced to something that is trivial.

Far-field diffraction resulting from elliptical apertures of different sizes is measured. Both normal and oblique angles of incidence on the apertures are investigated. Diffraction patterns from apertures with dimensions much greater than the wavelength agree with the existing theory. However, new diffraction patterns have been found when using an aperture of wavelength dimension.

The discovery of high-temperature superconductors (HTS) spawned many potential applications, including optical detectors. Realizing viable superconducting detectors requires achieving performance superior to competing and more mature semiconductor detector technologies, and quantum detector technologies in particular. We review why quantum detectors are inherently more sensitive than thermal or bolometric detectors. This sensitivity advantage suggests that for operation at cryogenic temperatures, we should be developing only quantum superconducting detectors. Accordingly, we introduce and describe the structure and the operation of a superconducting quantum detector with a superconducting quantum interference device (SQUID) readout circuit. The superconducting quantum detector, consisting of a superconducting loop, produces a photosignal in response to photoinduced changes in the superconducting condensate's kinetic inductance. The superconducting quantum detector is designed to operate only in the superconducting state and not in the resistive or transition states.

Cost-effective high-performance IR imaging cameras need affordable staring focal plane arrays (FPAs) that can operate effectively at temperatures compatible with inexpensive long-life coolers. We report on staring hybrid 128 x 128 and 256 x 256 Hg_1-xCd_xTe FPAs that have requisite yield, sensitivity, operability, and reliability at a medium-wavelength IR (MWIR) cutoff wavelength (λ_c ~4.6 μm at 180 K) and elevated operating temperatures. Mean 256 x 256 FPA noise-equivalent temperature differences (NEΔT) using broadband f/1.7 optics were 4.3, 7.7, and 55 mK at 120, 140, and 180 K, respectively. We extrapolate that camera NEΔT ≤ 0.02 K can be achieved at 190 K using optimized (λ_c of ~4.4 μm (180 K), a 3.4- to 4.2-μm bandpass filter, and f/1 optics. Because the CMOS multiplexers have a low-power dissipation and need little ancillary circuitry in the dewar, a viable thermoelectrically-cooled FPA technology is thus implied once the λ_c is optimized for MWIR imaging.

Single-layer and multilayer Si_0.7Ge_0.3/Si heterojunction internal photoemission (HIP) detectors with cutoff wavelengths out to ~23 μm have been demonstrated. Near-ideal thermionic emission dark current characteristics and photoresponse at wavelengths up to 20 μm were measured. The cutoff wavelength λ_c and emission coefficient C₁ of the HIP detectors were determined by the modified Fowler plot at the wavelength regime where the corresponding photon energies were smaller than the Fermi energy (~0.15 eV) of the degenerate Si_0.7Ge_0.3 layers. Similar optical and thermal potential barriers were obtained. The use of multiple Si_0.7Ge_0.3/Si layers in the stacked HIP detector structure resulted in a significantly increased emission coefficient C₁ compared to the single-layer HIP detectors due to an enhanced internal photoemission efficiency without the loss of IR absorption.

Results are presented for a modification of the spatial-harmonic-distortion test, suitable for nonlinearity measurement of single-element detectors, using a scanned-fringe technique and an electronic spectrum analyzer. Detector nonlinearity of less than 1% was measured at a wavelength of 10.6 μm. Previously, the smallest measured nonlinearity for the spatial-harmonic-distortion test was of the order of 3%, using a stationary-fringe technique and an 8-bit frame grabber.

The use of focal plane arrays in IR imaging systems is becoming increasingly important, but in the long-wave IR region the number of detector elements in the array is limited by the current state of technology, and this in turn restricts the available spatial resolution or field of view. An image processing scheme is described that is able to interpolate and enhance an image in a unified moving window operation. The image is first expanded to the required size by pixel replication and then processed so that the resulting spectrum approximates that of the original scene. A numerical method has been developed to calculate the interpolator and its use has been demonstrated in a computer simulation.

Reticle systems are considered to be the classical approach for estimating the position of a target in a considered field of view and are widely used in IR seekers and trackers. This type of seeker is most suitable for tracking airborne targets such as aircraft and missiles. With an image processing system, where the reticle pattern as well as the received radiation are described as digitized images of rather high resolution, simulations can easily be performed. The method can be used for the simulation of systems with spinning reticles as well as of systems utilizing fixed reticles with an off-axis rotating lens or mirror. In this manner any arbitrary reticle pattern can be studied and the pattern can be easily altered. Simulations were performed using a TERAGON® 4200 image processing system. The simulation method and the possible signal processing methods for different types of reticle systems are described. These systems are mainly used as seekers in surface-to-air or air-to-air missiles. The simulation model presented has been used in threat analysis and in studies on the effects of various types of countermeasures, e.g., flares and jammers. The result of the simulated reticle output signals has been validated by the use of different reticle seeker hardware.

A new control concept for instrument pointing, tracking, and vibration suppression is introduced based on zero annihilation periodic (ZAP) control. In ZAP control, the control gains vary periodically in time, in sharp contrast to conventional controllers whose control gains are fixed in time. The main advantage is that perfect "deadbeat" pointing, tracking, and vibration suppression can be achieved-even in the presence of flexible structural elements and noncolocated actuator and sensor hardware. The deadbeat response has clear advantages for optical instruments that must be held steady and precisely pointed during imaging. The ability of ZAP designs to effectively control noncolocated and nonminimum phase configurations opens up many new possibilities for high-performance instrument pointing, vibration damping, target tracking, and other advanced optics control applications.

Recent investigations of the atmospheric transmittance in the 15- to 25-μm spectral region have produced experimental results of the absorption of the water vapor continuum. From these results it is evident that the excessive absorption of the water vapor (the water continuum) cannot be explained simply by the theoretical continuum model based on the modified impact theory and the Van Vleck-Huber model of line broadening. The measured continuum absorption has spectral features similar to those roughly predicted by the water dimer theory. A quantitative agreement between the experimental results and the prediction according to the dimer theory was found.

The Infrared Space Observatory (ISO) camera (ISOCAM) is designed to map selected regions of the sky in the spectral region from 2.5 to 17 μm at various spatial and spectral resolutions and at high sensitivity (at >4 μm, <1 mJy/10 s in 200 5). It will make images, within the 3-arcmin field of view of the ISO telescope, with two 32 x 32 IR array detectors: an InSb charge injection device (CID) for the 2.5- to 5.5-μm range and a Si:Ga direct voltage readout for the 4- to 17-μm range. Four different pixel fields of view are available on each channel: 1.5, 3, 6, and 12 arcsec. The spectral range can be selected in each channel by a set of about 10 fixed bandpass filters (resolution from 2 to 100) and continuous variable filters (resolution ≈45); polarization measurements are possible as well. A very wide range of astrophysical problems can be tackled with ISOCAM. We present a brief description of the program planned by the ISOCAM team.

The camera of the Infrared Space Observatory (ISOCAM) was carefulfy tested and cafibrated from 1991 to 1992 in the dedicated facifity instafled at Institut d'Astrophysique Spatiafe at Orsay, France. A farge cryostat with high-precision instrumentation was devefoped to provide an operation environment cfose to the conditions expected in orbit. The large amount of data obtained allows a thorough investigation of the imaging, photometric, and spectral properties of ISOCAM in all its operation modes, leading to the definition of hopefully optimal observing procedures and data-processing algorithms.

A modified Twyman-Green interferometer is in use that makes possible wavefront testing of optical filters at any wavelength from 200 to 1100 nm. The use of mirrors for collimation and pupil imaging makes the instrument achromatic, and therefore the focus is fixed over the entire bandwidth. The beamsplitter and compensator plates are made of fused silica, and the detector is a UV-enhanced CCD TV camera. A tunable monochrometer with a broadband light source permits selection of any wavelength. Fringe distortion, even when the collimating mirror is spherical, is small enough to keep measurement errors within 0.1 wave peak-to-valley over the 2- x 2-in, aperture.

Recent breakthroughs in IR detector array and cryocooler technology have made it possible to convert the concepts of optimum, passive, IR sounding to a practical satellite-borne instrument: the Atmospheric Infrared Sounder (AIRS), a grating array IR spectrometer temperature sounder. AIRS, together with the Advanced Microwave Sounding Unit and the Microwave Humidity Sounder, will form a complementary sounding system for the Earth Observing System to be launched in the year 2000. The three instruments are expected to become the new operational sounding system for the National Oceanic and Atmospheric Administration.

As one of the four cornerstones of its scientific program Horizon 2000, the European Space Agency is studying a far-infrared and submillimeter space telescope (FIRST) that will open to astronomical observations the 100-μm to 1-mm spectral region, virtually unexplored with a subarcminute angular resolution. The 3- to 4-m-diam telescope will be diffraction limited at λ = 300 μm or less. The expected lifetime of this system in orbit is five years or more. The model payload is composed of two focal instruments. The multifrequency heterodyne spectrometer will give a very high spectral resolution in the 400- to 630-μm wavelength range and around 300 μm. The far-infrared receiver (FIR) will be an imaging spectrometer covering the 100- to 400-μm range with spectral resolutions ranging from 3 to 10⁴ obtained with a set of Fabry-Pérot interferometers and filters. Arrays of photoconductors and bolometers are cooled at 1.5 and 0.15 K, respectively, by an open-cycle ³He/⁴He dilution + Joule-Thomson expansion system developed to operate at zero gravity.

The concept of an actively deformable composite reflector is demonstrated and test verified. The objective of developing such a concept is to correct the low-order long-wavelength distortions in a space reflector. This is achieved by bonding a limited number of piezoelectric ceramic strips to the back face sheet of the reflector. A cryointerferometric test of a 0.5-m reflector with six groups of actuators was performed to verify the concept experimentally. Analysis of and test results for the development of an active deformable composite reflector panel are summarized.

To model the target acquisition capability of an electro-optic system operated by a human, one should take into account how the clutter in the scene affects the human performance. We consider one global and two local clutter metrics; we compare actual target acquisition results to those predicted by our metrics.

A model for the calculation of rms response nonuniformity in a direct-injection readout mercury cadmium telluride hybrid focal plane array is presented. The model includes contributions from detector material composition variations, processing-induced variations, threshold voltage variations in the input structure of the readout multiplexer, and transfer efficiency variations in the case of CCD-based readouts. The model is then used to analyze the role of the buffered direct-injection (BDI) circuit in controlling the nonuniformity. It is predicted that a minimum BDI gain of 10 is required to reduce the nonuniformity of long-wavelength infrared focal plane arrays (LWIR FPAs) to its minimum value limited by the compositional variations in detector material. This gain requirement is expected to go up with a reduction in the resistance-area product of photodiodes resulting from either imperfect processing or increased cutoff wavelength (λ_co > 9.43 μm).

We review the focusing properties of a general planesymmetric dispersive system containing a varied line-space (VLS) grating as its dispersive element. A VLS grating is one whose grooves are straight and parallel (when viewed looking down the grating normal) but unequally spaced. Aberration terms for VLS gratings are related to those for holographically produced gratings. The groove pattern is represented by a groove function, which, although less intuitive than the groove spacing, contains information that clearly corresponds to the imaging properties of the grating. The projection of the groove pattern onto the plane tangent to the grating at its center is shown to be useful in describing the focusing properties of the grating; consequently the imaging effects of the curvature of the grating blank itself can be considered separately.

We used isolated radioulnar bones (fused radial and ulnar bones) of subadult European moose collected in various environmentally polluted areas of Finland. The bones were radiographed and holographic interference pictures, for holographic nondestructive testing (HNDT), were produced by using small caudocranial bending forces. The cortical index values were measured in the diaphyses and samples were taken for mineral studies from the mandibles of the same animals. Results indicated that the bones obtained from the heavily polluted area showed biomechanical response comparable to the bones developed partially without mothers milk. Differences were also seen in morphometrical and radiological studies. The mineral contents studied did not differ significantly from randomly collected samples of the same age category. We therefore conclude that environmental factors may influence the bone matrix development.

A new acousto-optical holographic display device, which is expected to solve the display resolution and refresh problems that are common to available holographic displays, is proposed. The device is based on the reproduction of a hologram as a surface pattern using traveling surface acoustic waves (SAWs). There is an array of electrodes attached to the SAW device. An electrical signal applied to any one of these electrodes generates an acoustical wave propagating on the surface of the crystal. If signals are applied to all of the electrodes simultaneously, propagating waves from the electrodes superpose to form a time-varying SAW pattern on the crystal. This pattern forms the hologram at a specific time. The signals that should be applied to the electrodes are found through a mathematical inversion relation that is derived from the underlying physics. The inversion relation is checked by computer simulations. Simulations also show that the image quality of the proposed 3-D TV display is satisfactory.

A new method for noise factor determination in image intensifier tubes is described. The measurements made in second- and third-generation image intensifiers are in accordance with theory and experiments reported in the literature. This method allows one to carry out the measurement on the overall screen instead of the central area only.

A technique for optically implementing nonlinear gray-level transformations on images via a spatial light modulator (SLM) coupled to a computer image-processing system is presented. Both the exponential and the logarithmic functions are implemented. The only restriction in the choice of nonlinear transformations is that they must lie below the SLM's linear intensity output versus intensity input curve obtained with the SLM set to a full transmittance.

Several edge enhancement techniques are examined for improving the performance of binary phase-only filter pattern recognition devices. These include the linear techniques of blocking the lower Fourier orders and Laplacian filtering as well as the nonlinear techniques of phase extraction and phase binarization. The results indicate that the nonlinear techniques outperform the linear techniques when there is no distortion caused by aperturing the input information. However, when the input is apertured, both the linear and nonlinear techniques yield approximately the same peak-to-noise ratio.

Statistical detection theory is used to develop the structure and performance of quasioptimal detection processors for 3-D coherent laser radar range imagery. Generalized likelihood-ratio tests (GLRTs) and receiver operating characteristics (ROCs) are presented for a detection scenario involving a variety of unknown object and background parameters. A computationally efficient, hard-limiter matched-filter processor is shown to yield performance closely approximating that of the GLRT.

A novel lossless adaptive predictive coding scheme is proposed. The proposed coding scheme divides a coding process into two stages, one for the upper bit signal and another for the lower bit signal. The lower bit signal is adaptively encoded, after the upper bit signal is encoded, utilizing the upper bit signal spatial direction. In the proposed scheme, the predictor and entropy coders are designed to maximize their efficiency for encoding multilevel fingerprint images. The coding efficiency obtained by this scheme is much higher than conventional adaptive predictive coding, as far as multilevel fingerprint images are concerned, providing 9% entropy and 10% average code length reduction. Although the proposed method has been confirmed as being efficient only for encoding multilevel fingerprint images, it can be efficiently applied for encoding ordinary images, such as human faces, natural scenes, and so on, with minor modifications.

Atmospheric modulation transfer function (MTF) measurements carried out simultaneously in both thermal imaging atmospheric windows using both passive and actively heated targets are presented. Results indicate rather significant angular spatial frequency dependence of the MTF, in contradiction to the conventional approach, which assumes contrast transfer is atmospheric transmission only. A theoretical explanation is discussed, based on aerosol forward scattering and absorption effects, which is shown to be angular spatial frequency dependent and yields MTF results similar to those measured. This means that small details are blurred much more than large details by the atmosphere, thus also affecting target acquisition probabilities.

Commercially available CCD cameras adapted to standard television signal transmission have been analyzed for use in video speckle interferometry. Their relevant performance characteristics are described and the selection criteria for the different camera systems that are offered are discussed. It is shown that the requirements that are specific for video speckle interferometry often favor the use of cameras based on the interline transfer principle rather than frame transfer devices, especially when pulsed lasers are applied.

A grating interferometer for high-resolution displacement measurements is presented. To decrease interferometer sensitivity to inaccuracy of the grating guide, application of divergent light beam is proposed. The effect of the interferometer geometry on the direction and constant of fringes is analyzed. Comparison is drawn between results of applying plane and spherical wavefronts. The advantages of spherical wavefront application are theoretically evaluated and experimentally proved. An acceptable value of the angular instability of the grating illuminated by spherical wavefront is equal to about ±3 arcmin.

A prototype of a new stylus probe with an interferometric transducer for surface profiling is presented. Rotation of a measuring arm is transformed into electrical pulses by a multimode laser diode interferometer. A holographic grating laid on a cylindrical surface acts as a length master. The cylinder axis overlaps the rotation axis. Geometrical errors of the cylinder axis position relative to the rotation axis are analyzed. The influence of the offset of the measuring tip position relative to the measuring lever axis is emphasized. General characteristics of the device construction and metrological feasibilities of the prototype are presented. Resolution of 0.02 μm and nonlinearity error of ±0.05 μm in the measuring range of 7 mm are obtained.

We consider the mechanism of the speckle interlerometer output signal formation for biovibration measurements. Statistical characteristic analysis of the output signal amplitude is carried out. The factors creating amplitude-phase modulation of the output signal are investigated. The possibility of a differential homodyne speckle interferometer application in Tibetan pulse diagnostics is demonstrated.

The design of a six-stage liquid crystal tuned Lyot filter for operation at a wavelength of 532 nm is described. Design options including material type and factors affecting performance such as material quality are discussed. The proposed narrow-passband filter will have a bandwidth of less than 0.1 nm. To achieve the performance required of the filter, high transmittance polarizers have been developed.

A high-resolution moiré photography system is described in which a reflection phase grating attached to a specimen is imaged onto a reference grating to form moiré fringes. A tuned camera lens has been used to resolve grating frequencies of 150 lines mm^-1, giving a sensitivity of 6.7 μm fringe^-1 and to improve fringe contrast. The fringes are formed in real time and have been recorded with a high-speed image converter camera at framing rates of up to 10⁶ frames s^-1 with a field of view of up to 35 mm diameter. Sequences of fringe patterns have been analyzed by a 2-D Fourier transform method to obtain the dynamic displacement field with submicrometer accuracy. A field lens has been used to collimate the light incident on the specimen, and it is demonstrated that this makes the system insensitive to out-of-plane specimen motion, avoiding fictitious strains. An example is presented of the application of the technique to the measurement of the mode I stress intensity factor in a cracked polymethyl methacrylate (PMMA) plate under stress pulse loading.

Optical correlators with incoherent light offer many of the advantages of correlators with coherent light, yet they are not as sensitive to coherent noise and alignment errors. Moreover, sophisticated and complex spatial modulators are not needed. The correlation is performed by comparing the input scene with some predefined reference object that was recorded earlier in a single holographic filter. To increase the number of objects that can be simultaneously correlated, or, alternatively, allow rotation and scale changes of one object, we investigated how best to record holographically a multiplicity of different objects in one composite filter. We determined that for high SNR it is necessary to illuminate the recorded object with diffuse light. Results from computer simulation and laboratory experiments reveal that 16 different objects can be holographically recorded in one composite filter and then used for simultaneous correlation giving sharp correlation peak with good peak to background ratio. Alternatively, the range of rotation invariant correlation can be increased from a few degrees for a single filter to 160 deg for a composite filter of 16 objects.

A photonic circuit board is fabricated for potential application to interchip and interboard parallel optical links. The board comprises photolithographically patterned polymer optical waveguides on a conventional glass-epoxy electrical circuit board and a surface-mounted integrated circuit (IC) package that optically and electrically couples to an optoelectronic IC. The waveguide circuits include eight-channel arrays of straights, cross-throughs, curves, self-aligning interconnects to multifiber ribbon, and out-of-plane turning mirrors. A coherent, fused bundle of optical fibers couples light between 45-deg waveguide mirrors and a GaAs receiver array in the IC package. The fiber bundle is easily aligned to the mirrors and the receivers and is amenable to surface mounting and hermetic sealing. The waveguide-receiver-array board achieved error-free data rates up to 1.25 Gbits/s per channel, and modal noise was shown to be negligible.

We demonstrate an a-Si:H/su rface-stabilized ferroelectric liquid crystal (FLC) optically addressed spatial light modulator (SSFLC OASLM) incorporating a reflecting patterned metal mirror. The mirror, made using standard photolithographic techniques, has a three-layer structure containing two metal-matrix layers with an insulating layer between. Compared to a single metal-matrix layer, the gaps between the edges of pixels can be greatly reduced. With two metal-matrix layers, the principle advantages of a metal mirror over a dielectric stack reflector in FLC OASLMs are the metal layer's high capacitance and relatively achromatic reflectivity. A consequence of the metal-matrix pattern is that the effective reflectivity of the mirror is a function of the f/# of the read-beam-imaging optics. With f = 1/6, the theoretical maximum effective reflectivity is 0.91 R_metal where R_metal is the reflectivity of the metal. Our device has measured reflectivities as high as 0.52. The transmission coefficient of the mirror is measured to be as low as 0.017.

An experimental investigation was conducted of the feasibility in using thin film aluminum mirrors as components of an optical system designed to attenuate laser light. When the laser light is focused onto the surface of these mirrors, the aluminum film is vaporized. Using a unique optical geometry, the laser pulse is then returned to the initial damage spot to maximize attenuation. The amount of light that eventually emerges from the optical system is significantly reduced. Results indicate a maximum attenuation of incident light of more than a factor of 10⁴. Variations of the principal experiment were performed to optimize results. An explanation of some phenomena involved and possible improvements are discussed.

Optical, mechanical, and erosion-protective characteristics of boron and gallium phosphide are evaluated as single films and within antireflection multilayers. These coatings are shown to combine broadband infrared transmission with environmental durability, specifically in relation to abrasion resistance and elevated temperature performance up to 500°C. Rain erosion protection of all common IR optical materials is demonstrated from single water jet impact and whirling arm tests. Protective characteristics in relation to solid particle impact are described.

Ion figuring of optical components is a relatively new technology that can alleviate some of the problems associated with traditional contact polishing. Because the technique is noncontacting, edge distortions and rib structure print through do not occur. This investigation was aimed at determining the effect of ion figuring on surface roughness of previously polished or ductile ground ceramic optical samples. This is the first step in research directed toward the combination of a prefinishing process (ductile grinding or polishing) with ion figuring to produce finished ceramic mirrors. Multiple, chemical vapor deposited silicon carbide (CVD SiC) samples were polished or ductile ground to specular or near-specular roughness. These samples were then characterized to determine topographic surface information. The surface evaluation consisted of stylus profilometry, interferometry, and optical and scanning electron microscopy. The surfaces were then ion machined to depths from 0 to 5 μm. The finished surfaces were characterized to evaluate the effects of the ion-machining process with respect to the previous processing methods and the preexisting subsurface damage. This study provides some of the information required to effectively utilize a combined ductile grinding or polishing with ion machining as a procedure for figuring optical components.

A flow visualization system is described that provides 3-D snapshots of instantaneous flow structures by rapidly acquiring successive image planes from within the flow. For each 3-D snapshot, up to 20 planar images are collected at the rate of 10,000,000 images/s. Flow visualization at successive planes is by planar particulate scattering or planar laser-induced fluorescence using a laser sheet rapidly swept once across the flow by a 30,000-rpm polygon scanner. Variations in the laser sheet intensity are monitored by placing an on-edge image of the sheet on the periphery of every planar flow image. The camera is a high-speed image converter coupled by fiber optics to a slow-scan CCD camera. The fiber optics provide 2.5 times better light gathering efficiency than lens coupling without reducing the available spatial resolution. The overall measurement technique shows excellent potential as demonstrated with a selection of 3-D measurements.

Based on a pinhole camera model, camera model equations that account for the radial lens distortion are used to map three-dimensional (3-D) world coordinates to two-dimensional (2-D) computer image coordinates. Using two cameras to form a stereo vision, the 3-D information can be obtained. It is demonstrated that such stereo imaging systems can be used to measure the 3-D displacement field around the crack tip of a fracture specimen. To compare with the available 2-D theory of fracture mechanics, the measured displacement fields expressed in the world coordinates are converted, through coordinate transformations, to the displacement fields expressed in specimen crack tip coordinates. By using a smoothing technique, the in-plane displacement components are smoothed and the total strains are obtained. Rigid body motion is eliminated from the smoothed in-plane displacement components and unsmoothed out-of-plane displacement. Compared with the theoretical elastic-plastic field at a crack tip, the results appear to be consistent with expected trends, which indicates that the stereo imaging system is a viable tool for the 3-D deformation analysis of fracture specimens.

This pdf contains the communication on "Spatial filtering by holographic optical elements."