The power dissipation for large-area high-speed charge-coupled device (CCD) arrays is analyzed. The four mechanisms responsible for power dissipation in a CCD are carrier lift and friction within the CCD channel, load currents through the on-chip output amplifiers, capacitive charge and discharge currents through the epitaxial layer or substrate, and currents flowing through the polysilicon clock register electrodes. These processes are easily calculated for small or slow devices. For large high-speed devices, on the other hand, the conventional analytical techniques are inadequate. The large device area leads to high resistive-capacitive constants in the parallel clock registers, causing clock pulses to degrade in shape as they travel along a clock electrode. A distributed system analysis based on lumped circuit parameters for each pixel is necessary to calculate the current and thus the power dissipated at each point on the surface of the COD. A computer simulation of a standard frame-transfer CCD was performed using SPICE software. The results of the simulation indicate a much lower power dissipation than previously assumed and also point out various problems with conventional device architectures for large-area high-speed CCDs.

Four types of transparent electrically conductive filters commonly used for a variety of applications are (1) a free-standing conducting inductive mesh, (2) a substrate coated with a conducting inductive mesh, (3) a conducting substrate, and (4) a substrate coated with a continuous conducting coating. Each can be designed to provide high visible, near infrared, and/or far infrared transmittance and high radio frequency (rf) and/or microwave attenuation. Theoretical models developed at Itek to predict the optical, rf, and electrical properties of each type of filter as a function of conductivity, relaxation time, electronic mobility, thickness, mesh period, and mesh line width are described. The temperature and angle of incidence dependence are also discussed. Simple expressions for predicting the performance of these four types of filters at both optical and rf frequencies provide both insight and useful starting designs.

A new method for measuring spatial coherence in which optical fibers replace the traditional slits in Young's experiment is developed. The beam is sampled by the fiber ends, and the phase difference that causes the fringes is introduced by translating one fiber along its long axis by a few wavelengths. The technique has several advantages over the slit method: The fibers can be moved to arbitrarily chosen points, as opposed to removing and replacing sets of slits, giving greater flexibility in measurement as well as easing alignment problems. This method also does not rely on diffraction, and therefore enables measurement of much dimmer beams and dimmer points on beams. One can take as many data points as desired, thus tracing the shape of the coherence function accurately, whereas with slits one is restricted to the number of slit sets available. Finally, one does not have to assume beam symmetry. Although the magnitude of the spatial coherence function is not accurately reproduced, this can be corrected for, and the ability to reproduce the shape of the beam remains a useful measurement tool. We report, in particular, measurements of spatial coherence of a single-stripe diode laser, as well as a HeNe laser.

The blockwise transform edge detection method consists of segmenting the image into a number of small blocks, with neighboring blocks overlapping slightly; transforming each block by a fast transform; multiplying the transform coefficients of each block by a bandpass mask; inverse transforming the modified transform coefficients; detecting zero crossings and deciding an edge according to a thresholding scheme. Simulation results with the blockwise transform edge detection method in comparison to other bandpass filtering techniques indicate that higher performance is achieved with the new technique on the average. This is believed to result from the use of blockwise generalized filtering with appropriate transforms rather than linear convolutional filtering, as in other bandpass filtering techniques. The new method also has less computational complexity, and is well-suited to parallel processing.

A new quality metric for evaluating edges detected by digital image processing algorithms is presented. The metric is a weighted sum of measures of edge continuity, smoothness, thinness, localization, detection, nd noisiness. Through a training process, we can design weights that optimize the metric for different users and applications. We have used the metric to compare the results of ten edge detectors when applied to edges degraded by varying degrees of blur and varying degrees and types of noise. As expected, the more optimum Laplacian-of-Gaussians (LoG) filter and Haralick's second derivative method outperform the simpler gradient detectors. At high SNR, Haralick's method is the best choice, although it exhibits a sudden drop in performance at lower SNRs. The LoG filter's performance degrades almost linearly with SNR and maintains a reasonably high level at lower SNRs. The same relative performances are observed as blur is varied. For most of the detectors tested, performance drops with increasing noise correlation. Noise correlated in the same direction as the edge is the most destructive of the noise types tested.

Mercury-cadmium-telluride (HgCdTe) infrared focal-plane arrays (IRFPAs) based on photodiodes and photoconductors are compared. The comparison includes readout structures for both types of detectors, analysis of the SNR of the various configurations, and the noise requirements of the coupling devices. The feasibility of IRFPAs based on linear arrays of photoconductors is discussed, and the significance of such a novel IRFPA is presented.

A number of points concerning display holography and a method of so-called pseudo-deep holograms are reviewed. The possibility of using the angular selectivity property of pseudo-deep holograms for 3-D image formation is investigated. The optical system characterized by the presence of an inclined hologram and a horizontal slit that filters the beams from an object and those from an extended reference source is considered. It is shown that the hologram recorded with the help of such a system is capable of reconstructing 3-D images while illuminated by an extended light source characterized by a distribution of phases different from those of the reference source. This record is referred to as a selectogram. The resolution of the image reconstructed by the selectogram is analyzed. The results of the recording and reconstruction of selectograms of a special 3-D arbitrary diffuse object are considered. It is noted that it is possible to record a selectogram using the radiation scattered by the object as the reference wave.

We propose a new technique to calculate the phase distribution over a fringe pattern. A digital processing system is used to digitize and analyze several holographic interferograms. Starting from a well-known tracking algorithm, which has been optimized, interferometric fringes are codified and then, by taking advantage of the intensity distribution over the interferogram, the phase associated with every pixel is calculated. We conclude that this technique can be applied to all kinds of fringe patterns obtained by the different techniques of holographic interferometry.

We use flicker to study the sensitivity of the human visual system to color differences of pairs of alternating color stimuli and obtain dynamic color-discrimination ellipses. Our results are compatible with the present theory on color-discrimination ellipses and color mixture.

We demonstrate the application of optical amplifiers and polymer-dispersed liquid-crystal (PDLC) shutters to electrically reconfigurable fiber optic delay line signal processors. Two 8-tap finite impulse response (FIR) fiber filter modules were fabricated. These two modules can also be interconnected in a cascaded or parallel configuration to implement a 16-tap fiber FIR filter. An erbium-doped fiber amplifier with a peak gain of about 20 dB was used to compensate for the large optical losses involved in the filters due to the large tap numbers. The relatively inexpensive PDLCs were used to realize electrically reconfigurable analog tap weights. The individual fiber filters were then evaluated for their impulse and frequency responses. The fabricated filters used single-mode fibers and fiber components and were polarization independent to within 0.5 dB. The sampling frequency was about 200 MHz, which can easily be upgraded into the gigahertz range. The tapping extinction ratio was about 13 dB with subkilohertz tunability speed. The amplified spontaneous emission noise can limit the filtering performance unless appropriate spectral filtering is included before detection. These optically amplified electrically reconfigurable fiber signal processors have the potential to lead to the realization of complex programmable and adaptive optical systems.

Effective implementation of image processing algorithms for enhancement and restoration often assumes that the images are degraded by known statistical noise. Depending on the application, the type of noise present may vary. The noise distributions that are commonly encountered in image processing are the Gaussian, Rayleigh, negative exponential, and gamma. Typically, when computer-generated noise images are used for algorithm development they are spatially uncorrelated. It is the purpose of this paper to present various types of computer-generated two-dimensional correlated and uncorrelated noise images along with suggestions of several applications.

We describe the development of a 3-D statistically nonstationary earthlimb stochastic structure representation and its application to earthlimb infrared (IR) background structure simulations. An earthlimb viewing geometry is defined by the location of both observer and source in space. The line-of-sight tangent altitude is the minimum altitude of the trajectory, defined at the earth-centered perpendicular to the line of sight. The stochastic structure overlay is constructed from a 2-D diagonal cut through a 3-D matrix of correlated Gaussian deviates; each plane of the matrix represents a statistically constant representation of the 2-D correlation lengths at a given altitude above the earth. Each matrix plane is generated using successive 2-D fast Fourier transform routines that have empirical values of vertical and horizontal correlation lengths as input. The total deviate variance versus altitude is then scaled from empirical measurements of fluctuations in atmospheric density, temperature, and/or emissivity. These structure generators are used both as perturbations on input atmospheric data to IR radiance codes and as high-resolution overlays to earthlimb lR mean-radiance 2-D scenes. An IR structured scene realization and scene validation analysis is presented.

We discuss errors and uncertainties in calibrating infrared radiometers to measure the temperature of clouds in the earth's atmosphere. Many of the points we make apply to any radiometric temperature-sensing application where the target temperature is less than the ambient temperature. Dominant uncertainties and errors are due to ambient radiance reflected from the blackbody-simulator source, thermal fluctuations in the radiometer, and imprecise voltage measurements. Our improved technique removes, reduces, or accounts for these errors and uncertainties. The resulting calibration uncertainty is ± 0.8°C for a radiometer with a 10-v output range. We verified this accuracy by comparing cloud-base temperatures measured by ground-based IR radiometers, in situ (radiosonde) sensors, and other remote sensors on the ground and on satellites. We made these comparisons for spatially uniform blackbody clouds that filled the field of view of our ground-based radiometers.

Waveguide devices have been produced comprising germania-doped silica cores and silica claddings on silica substrates, using microwave plasma assisted chemical vapor deposition and reactive-ion etching. This all-silica structure offers the maximum compatibility between fibers and waveguides in terms of both optical and physical properties. The all-silica philosophy is extended to the use of laser-cut silica V grooves in the construction of input/output fiber arrays. An important feature of this approach is that it enables fiber/waveguide interfacing by CO₂ laser welding of the respective silica blocks without the introduction of any significant additional loss. Both the waveguide fabrication and laser-welding processes are well suited to large-scale production of low-cost components.

The contributions of geometry dispersion, film dispersion, and substrate dispersion of a homogeneous strip waveguide to its total dispersion are discussed individually by means of the decomposition method. Universal formulas and curves are also given in terms of normalized dispersion parameters.

Optical interference filters that have continuously modulated refractive indices throughout their thicknesses (rugates) are designed and fabricated using a microwave plasma-assisted chemical vapor deposition technique. Examples are shown of single- and double-response narrowband reflection filters made from variable composition silicon oxynitride deposited on silica. Reflectivities in excess of 99.9% have been realized with bandwidths in the 2 to 10% range.

Combustion in a propane/air burner is analyzed by means of imaging molecular constituents present in the plane of a laser light sheet and by recording the local refractive index distribution with a shearing interlerometer. The pulsed illumination with a UV light sheet yields an instantaneous image of the distribution of propane by elastic scattering, and of OH and CH radicals by laser-induced fluorescence. Spectral separation allows simultaneous and separate imaging of elastic and inelastic scattering, thus providing a means to analyze the spatial inter-relation of propane, CH, and OH. In a fuel-rich flame, laser-induced incandescence of soot particles is observed. Calibration of propane concentration levels is achieved by the refractive index measurements, which also help to estimate flame temperatures.

A real-time displacement measurement system using a sinusoidal phase modulating interferometer is proposed and demonstrated. The system utilizes the frequency components of the interference signal to detect the desired phase or the displacement in real time. The real-time phase detector is constructed with simple electronic circuits. The system also has a feedback loop to eliminate external disturbance. It is easy to measure the relatively large displacement that exceeds a half wavelength of the light because the compensation circuit is incorporated into the system to obtain continuous phase outside the region from 0 to 2π.

The spot size measurement of a Gaussian laser beam is extended to wide dynamic range through the formulation of a novel exponential grating. It is found that the exponential grating provides accurate measurement for both extremely small and large Gaussian beam diameters and proves to be superior to Ronchi, triangular, and sinusoidal rulings.

A hybrid grid-moiré method for measuring strain is presented. The method employs a line grating attached to a plane surface of the specimen. Images of undeformed and deformed gratings are acquired and stored using a video system and a frame digitizer. The magnification is adjusted to record 10 to 30 lines of a grating with a pitch of 4 μm over the frame width. A fast Fourier transform (FFT) of the intensity data gives a close estimate of the spatial line frequency. A moiré function is formed as the product of a reference sine function and the image intensity-position function. FFTs of the moiré function yield an improved estimate of the line frequency. The line frequency is related to an average strain over the recorded image. The method was demonstrated at a magnification of 1500 x with a grating pitch p_s = 4 μm. The accuracy achieved depended on strain. For strains in excess of 0.1%, the errors were usually less than 5% when compared to strain gauge results. For plastic strains, in excess of 1%, the differences between two comparable measurements were less than 3%.

A three-layer neural network implemented by optical and electronic techniques for pattern recognition is described. The principle, the architecture, and the experimental results of the hardware system for pattern recognition are presented. A multiplex hologram is used as the interconnections between neurons of the input layer and the middle layer. The winner-take-all (WTA) operation of the intermediate layer is performed by an electronic circuit. The interconnection between neurons of the middle layer and the output layer is simply implemented with an optical transparency. By changing the patterns on the transparency, this system can be used for both auto- or heteroassociative memories. In addition to noise and partial hindrance tolerances, it also has illuminance invariance because of the WTA layer of the system.

Using luminaires based on the compound elliptical and compound hyperbolic concentrators of nonimaging optics, extremely uniform illuminance on distant targets can be achieved by utilizing pair-overlap flux maps. Such luminaires would also ensure maximal lighting efficiency (no trapped radiation), excellent angular (glare) control of radiation, and minimal sensitivity to errors in luminaire spacing.

We have investigated the nonlinear optical mechanisms responsible for optical limiting of both picosecond and nanosecond 532-nm optical pulses in the organometallic compound cyclopentadienyliron carbonyl tetramer (King's complex). For fluences below ~200 mJ/cm², picosecond pump-probe measurements in solutions of the King's complex reveal a prompt reverse saturable absorption (RSA) that recovers with a time constant of 120 ps. We attribute this RSA to excited-state absorption within the singlet system of the King's complex, and we demonstrate that the RSA is completely characterized by a simple three-level model. We find, however, that the material parameters extracted from these picosecond measurements cannot account for the strong optical limiting previously observed in identical solutions of this compound using nanosecond excitation at higher fluences. Picosecond measurements at fluences greater than 200 mJ/cm² reveal the onset of an additional loss mechanism that appears ~1 ns after excitation. The magnitude of this loss depends on both the laser repetition rate and the solvent, indicating that the loss is not directly related to the intrinsic properties of the King's complex but is most likely thermal in origin. Using nanosecond excitation pulses, we have performed angularly resolved transmission and reflection measurements, which reveal strong forward- and backward-induced scattering at these fluences. Furthermore, when the King's complex is incorporated in a solid host, we observe negligible induced scatter and the response is completely described by the singlet parameters extracted from the picosecond measurements. These observations indicate that the nanosecond optical limiter response of solutions of King's complex is dominated by thermally induced scattering.

The transient response of least-squares-based multiwavelength pyrometers to rapid temperature changes is modeled. Results are presented for a one-parameter and a four-parameter linear least-squares pyrometer with 31 wavelengths from 500 to 800 nm. The one-parameter response is very similar to the transient response of a two-color pyrometer. The response is more rapid than expected when the target temperature is rising and slower than expected when the target temperature is falling. The four-parameter pyrometer can overfit and indicate erroneous temperatures.

We present a technique for interferometrically testing aspheric surfaces without the use of compensating elements. The method consists of recording successive overlapping phase maps from a set of annular interferograms of an aspheric surface, obtained using a conventional phase-shifting interferometer and a micropositioning translator stage. These maps are then sewn together with a suitable algorithm we developed, and the whole surface error is recovered. Experimental results are shown to be in good agreement with the null lens test performed for comparison.

A new birefringent common-path interterometer with an automatic data-processing system is presented fortesting large convex spherical surfaces. With this interferometer, it is easier to eliminate system errors and to achieve quite accurate measurement. The precision of the system is about λ/50 with an accuracy of about λ/10.

For the Ronchi test, the phase of the Ronchigram can be retrieved very accurately with the aid of a phase-reduction algorithm originally designed for phase-shift interferometry. The simplest algorithm to reproduce the test wave front is the integration of the Ronchigram phase. However, the accuracy of integration depends heavily on the grating spacing. The effects of grating spacing is discussed using actual test results and mathematical simulation. The results indicate that coarse gratings, which benefited from the accuracy of phase-reduction techniques and qualified for integration in retrieving wave fronts, may be the better choice for the Ronchi test.

A simple method has been used to boresight an IR imager. The mechanical axis is defined by a plane parallel mirror attached to a reference mechanical flange. The difference between the optical and mechanical axis was measured by a dual-band modified autocollimator. The boresight is achieved by a vertical lateral movement of the lens relative to the detector for vertical orientation and by a zero offset of the scanning mirror direction for horizontal orientation. The accuracy of the measured angles depends on the focal length of the autocollimator optics and the electronics for measuring the image position by the video output of the thermal imager.

In a previous paper in 1992, we presented a flow visualization technique based on deflection mapping. We now present a new method for the interpretation of deflection mapping of small angles obtained by Ronchi gratings. In this method, a Fourier transform is operated on the transmission pattern of the Ronchi grating and the individual sine terms are then obtained by applying an inverse Fourier transform of each harmonic separately. The extraction of the deflection angle profile from the sine term is straightforward. The theory of Fourier transform deflection mapping is presented and the methodology for determining the deflection angle profile is discussed. The method was validated by measurements of transmission patterns of long-focal-length (4-m) lenses. A parallel light beam was passed through a lens, then traversed a Ronchi grating, and the transmission pattern was measured at a distance from the grating. The transmission patterns were measured by two methods (1) scanning with a photomultiplier tube over the grating and (2) using a linear 4096-photodiode array. A fast Fourier transform (FFT) code was used to isolate the various harmonics and the deflection angle profile was determined from the first harmonic. The focal lengths were determined by the present method with precision better than 99.999%, whereas the focal length was within 1% of the value provided by the manufacturer. Special emphasis was given to evaluating the errors associated with the technique and detectibility limits were given.

A typical reticle tracker includes a single frequency-modulated (FM) reticle to encode target location. This system is not useful for tracking multiple or large targets. A system, using two FM frequency-versus-radius reticles, is presented that allows tracking of these type targets. Results of computer simulations for both the single- and double-reticle tracking systems are presented and their performance is compared and contrasted.

The values of the quantum yield, its spectral dependence, and the external quantum efficiency of GaAlAs/GaAs transmission-mode photocathodes used in image intensification systems are computed based on the parameters of the internal quantum efficiency, the diffusion length, and the normalized surface recombination velocity. The variations for these parameters obtained are shown on the three orthogonal axes. Possible explanations and further lines of the investigation are also discussed.

The general case of the sign of the size-dependent strain in defect-free cubic microcrystals bounded by {100} faces is examined. Whereas the (F/L)^1/2 dependence of strain is confirmed, F being the surface free energy per unit area, and L the microcrystal edge length, the sign of the variation of lattice constant is found to be determined by the sign of surface tension. Some experimental and theoretical results are discussed.

Many robotic operations, e.g. , mapping, scanning, feature following, etc. , require accurate surface following of arbitrary targets. A versatile surface-following and mapping system using a very high precision "eye-in-hand" proximity sensor, designed to provide hardware, software, and application independence, modular development, and upward expandability, is presented. These goals are met by a full a priori specification of the hardware and software interfaces; a modular system architecture; and a hierarchical surface-data analysis method, permitting application-specific tuning at each conceptual level of topological abstraction. This surface-following system was fully designed and implemented independent of any specific robotic host and then successfully integrated with and demonstrated on two very dissimilar real-time robotic systems. Presented in detail are the hardware and software architectures, the mapping and terrain forecasting algorithms, and the applications' environments and task requirements.

Spatial characteristics of a thermal lens induced by a Gaussian beam in an absorbing liquid are studied using a nonlinear geometrical optics approximation. We show that the thermal lens in the vicinity of the sample cell is a focal point of light and a ring-of-light foci located in a plane displaced with respect to the focal point. The distance between the focal point and the ring-of-light foci decreases when the light power is increased. Experiments performed on an ethanol solution of iodine confirm the existence of this microscopic structure. The distance between the focal point and the ring of foci was measured for samples of different absorption coefficients. A good agreement between theory and experiment was obtained.