We announced a new section of Optical Engineering called Communications in the April issue of 1992 Vol. 31, No. 4􀀀. I wrote in the editorial of that issue about the concept behind this new section. The purpose of this new section is to publish short communications on original work, comments on recent papers, and errata. It will be my responsibility to de-termine what material is accepted for publication and, when necessary, I will seek outside reviewers’ assis-tance. This is not meant to be a ‘‘letter to the editor’’ section where writers comment on issues of the day, whether political or scientific. Contributions must be technical in nature and add something positive to our further understanding of our field of optical science and engineering.

We report a novel method to test and estimate collimation and beam alignment using a liquid-crystal TV (LCTV). In this technique, a lenslet array pattern generated using a computer program is ‘‘written’’ (transferred) as a video signal to the LCTV using an image-processing system. The image formed by this array at a suitable distance is grabbed using a CCD camera and processed using a centroid-finding algorithm. Experimentally, deviations from collimation and beam alignment as small as 10 µm have been detected. Using a LCTV, it is possible to obtain near-perfect collimation and beam alignment. This technique will be used to achieve initial system collimation and/or alignment and as an aid in automating optical systems that are required to maintain constant collimation and beam alignment.

The high-precision inspection and alignment algorithms of surface mounting devices (SMDs) using fuzzy morphology are proposed. After camera calibration is performed, the proposed inspection algorithm detects SMD lead defects such as bending and breaking using the ruler generated by fuzzy morphology. The SMD having no defects is tested to determine whether or not it is mounted correctly in the specified position. The proposed alignment algorithm accurately computes orientation and position using subpixel interpolation. It consists of two parts: the preprocessing and main processing parts, in which corner points and coarse orientation of an SMD are detected and interpolation is used to obtain final parameters with subpixel accuracy. The computer simulation with several sets of test images shows that the proposed algorithms give accurate parameters, and they can be applied to fast and accurate automatic surface mounting systems.

Modern avionics equipment, such as superresolution direction-finding systems (frequency band 0.3 to 3 MHz), now require resolutions greater than 15 bits. Oversampled analog-to-digital converter architectures offer a means of exchanging resolution in time for that in amplitude and represent an attractive approach to implementing precision converters without the need for complex precision analog circuits. Using oversampling techniques based on sigma-delta (??) modulation, a convenient trade-off exists between sampling rate and resolution. One of the major advantages of integrated optics is the capability to efficiently couple wideband signals into the optical domain. Typically, ?? processors require simple and relatively low precision analog components and thus are well suited to integrated optical implementations. The current ?? methodology is reviewed and the design of the single-bit, integrated optical ?? modulator is presented. Simulation results for both first- and second-order architectures are presented by evaluating the transfer characteristics numerically. A 16-bit f 0=1 MHz design (oversampling ratio of 132) is also quantified.

The author describes a novel design of neural networks for lossless data compression. The proposed neural network establishes an efficient dictionary by storing two symbols into each neuron and interconnecting those neurons that match a number of consecutive input strings. After an operation of experience-based competitive learning, a number of input strings can be matched by winning neurons in the network. Variable-length codes are then designed to encode the location of the first neuron, possible interconnections, and the number of matched neurons to achieve data compression. For unsuccessfully matched input strings, a literal code is constructed that contains an overhead code identifying the length of the literal code and the original codes of those unsuccessful strings. Extensive experiments show that the proposed neural network achieves very competitive compression performance in comparison with a few typical existing data compression algorithms. This also opens a new area for the application of neural networks to lossless data compression, where massive parallel processing and powerful learning capability can be utilized to develop high-performance algorithms and new techniques.

The polarization dynamics of vector solitons are studied in an elliptically low birefringent optical fiber for the first time within the framework of a model described by a system of coupled nonlinear Schro¨dinger equations with all the oscillating terms in the coupling between two copropagating modes. A condition for their stability has been obtained. Our analysis is based upon Hamiltonian perturbation theory.

Computing applications continue to put increasing demands on secondary storage systems. These demands can be met by optical solutions such as volume holographic memories, which promise to provide the high storage capacities and data rates required by many input/ output intensive applications. Holographic memories appear especially suited for large database storage because of their page format and fast associative recall ability. A volume holographic database system that is being implemented in our laboratory is presented. The discussion includes a description of the optical system as well as the development and analysis of an error-correcting code suitable for the application. Also presented are experimental results in which good associative addressability is achieved for search arguments as small as 1/30 of a page.

In this paper we present a method for segmenting document page images into text and nontext regions. The underlying assumption made by this approach is that the two regions can be viewed as different textures. We do not use any a priori knowledge of the document format. A convolution-based method is used to generate the texture feature images. The coefficients of the convolution masks are obtained using a single-layer artificial neural network that generates eigenvectors of the correlation matrix of the input data. The coefficients of these masks have been ‘‘learned’’ from examples of the document images and have a potential of being considerably more powerful than masks with preset coefficients. A thresholding scheme based on a measure of entropy is used to segment the feature images into the homogeneous regions.

A new adaptive algorithm for image interpolation with perceptual edge enhancement is proposed. Here, perceptual means that edges are enhanced and interpolated in a visually pleasing way. Each pixel neighborhood is classified into one of three categories (constant, oriented, or irregular). In each case, an optimal interpolation technique finds the missing pixels without generating unpleasant artifacts such as aliasing or ringing effects. Furthermore, a quadratic Volterra filter is employed to extract perceptually important details from the original image, which are then used to improve the overall sharpness and contrast. Both qualitative and quantitative simulation results clearly show the superiority of our method over standard low-pass interpolation algorithms such as bilinear, diamond-filter, or B-spline interpolation.

We evaluate the performance of a feature-preserving filtering algorithm over a range of images corrupted by typical additive random noise against three common spatial filter algorithms: median, sigma and averaging. The concept of the new algorithm is based on a corruptedpixel identification methodology over a variable subimage size. Rather than processing every pixel indiscriminately in a digital image, this corrupted-pixel identification algorithm interrogates the image in variable-sized subimage regions to determine which are the corrupted pixels and which are not. As a result, only the corrupted pixels are being filtered, whereas the uncorrupted pixels are untouched. Extensive evaluation of the algorithm over a large number of noisy images shows that the corrupted-pixel identification algorithm exhibits three major characteristics. First, its ability in removing additive random noise is better visually (subjective) and has the smallest mean-square errors (objective) in all cases compared with the median filter, averaging filter and sigma filter. Second, the effect of smoothing introduced by the new filter is minimal. In other words, most edge and line sharpness is preserved. Third, the corrupted-pixel identification algorithm is consistently faster than the median and sigma filters in all our test cases.

We describe the development and the implementation of a new nonlinear image processing methodology to detect dim targets in infrared images. The methodology is based on mathematical morphology. In a first phase, the image is filtered in order to enhance positively contrasted, isotropic objects with a specified size. Then, a simple motion analysis is carried out, and finally, detection is performed by thresholding. A complete study of the algorithm performance on NATO test sequences is presented. A comparison is made with the Holmes filter and the scale-subtraction filter, a wavelet-based method.

The answers to two common questions are attempted. The first is the mathematical representation of the visibility characteristics of a retroreflector. The second question concerns the construction of the necessary viewing criterion under highly nonuniform illumination. The problem of viewing a retroreflective object is discussed from the general viewpoint of the optical reciprocity theorem. Nonuniform illumination leads to peculiarities in calculating the contrast because the absence of any reference image region that usually serves as a background. For this reason, an optical sensor focuses its attention on the maximal brightness gradient of an image to compare the radiances of two adjacent resolution elements. In such a case, the valid signal is shown to be determined only by the radiance of direct light from a dummy source having an angular pattern of radiation that is the same as the angular pattern of the real receiver sensitivity, whereas optical interferences are obtained by taking multiply scattered light into account. The calculations are made under the small-angle diffusion approximation with the direct light isolated. One interesting example of observing retroreflective and Lambertian objects is presented. In particular, the case of viewing a retroreflector on a Lambertian background with the same albedos is considered. Here, the observation is possible only due to the different angular patterns of reflection.

A cosine wave encoded joint transform correlation is proposed that can deliver excellent correlation performance in the presence of nonoverlapping noise or additive noise, and guarantee that the optical gain is less than unity—facilitating implementation. The method proposed uses three CCDs to capture the power spectra of the joint image, the reference-only image, and the input scene, and then extracts the cosine wave information of the joint power spectrum using simple subtraction and division operations. The extracted cosine wave information is encoded onto the input SLM of the second optical setup, resulting in a high optical efficiency. The cosine wave encoded joint transform correlation is found to be similar to even-function correlations between the phase-only Fourier-plane information of the reference and input images, which resembles the correlation of the phase-only input scene information with the phase-only filter using a reference image in the VanderLugt correlation arrangement. Computer simulations show that the cosine wave encoded joint transform correlation outperforms binary joint transform correlation when the targets are embedded in either nonoverlapping noise or scenes with additive noise; performance is maintained even when the target has a very low light contrast ratio in a severely noise-corrupted input scene.

A PC-based, 3-D surface profile and displacement measurement system capable of micron-level accuracy using moderately priced off-the-shelf equipment has been developed. For use in field applications, a simplified calibration process using precision grids and camera translations is developed. An improved image correlation process is developed which corrects for perspective distortions due to viewpoint differences between the two cameras. The accuracy of the system was assessed experimentally and results expressed using several different error measures, including a new error measure proposed by the authors. The accuracy for both the profile and displacement measurement systems was established through a series of profile and translation tests. The baseline tests confirmed that the measurement system is capable of highly accurate full-field measurements. The system was also used successfully to measure both the bending of a clamped circular plate under pressure loading and the local buckling which occurs during tension loading of a cracked plate.

The plateau equalization algorithm for display of IR images is defined to include histogram equalization and histogram projection as special cases. A maximum gain parameter is defined and surveyed for a large number of images, determining the useful range of plateau values. Analysis of hardware experience with histogram projection explains why it produces intensity jitter in very low contrast scenes. Detailed flow charts for hardware implementation of plateau equalization are provided.

The performance of a long-baseline optical stellar interferometer is greatly enhanced if the instantaneous atmospheric delay ?(t ) can be tracked to within a fraction of a wavelength, permitting coherent integration of the optical correlation (fringe visibility). Real-time fringe tracking involves a control system that servos a rapidly responding pathlength compensator in real time. However, precise delay tracking can be achieved at somewhat lower signal levels by employing an off-line delaytracking system, in which the raw data measured by the interferometer are stored for subsequent analysis. Then the estimate of ? at time t is based on data collected both before and after time ? . An optimum delaytracking algorithm embraces the a priori statistics of the atmospheric delay process. Rather than simply estimating t at a point in time, a superior estimate of ? will be obtained by comparing all possible functions ?(t ) over a time period. Using Bayes’s theorem, the a posteriori probability density of any t(t ) function can be determined. An algorithm is developed that determines one or more functions that maximize that probability. Even the ambiguous estimates that result at lower signal levels can be employed for the coherent integration of optical correlation.

A large-aperture high-accuracy phase-shifting digital flat interferometer that is a combination of optics, mechanics, electricity, and algorithm is described. The aperture size is 250 mm and a liquid surface is used as an absolute flat to eliminate system error. The accuracy is better than ?/50 (?50.6328 µm, peak-to-valley value). The tested aperture can be enlarged to 500 mm. This interferometer has been used as an optical flat standard instrument for China. The optical interferometer, phase shifter, and calibration of precision are described.

A modification of a dual-beam illumination arrangement for electronic speckle pattern interferometry (ESPI) that leads to a twofold increase in the sensitivity to in-plane displacements is described. For each of the two illumination beams, the scattered light is observed along the direction of illumination, and these two components of the scattered light are then combined. The error in perspective caused by oblique observation can be almost corrected for by using two right-angle prisms in front of the object being studied. Experimental results including phasestepping are presented and compared with the results predicted by theory.

The paper presents a distance meter for noncontact measurement over extended ranges based on the frequency-modulated continuous-wave (FMCW) technique. A tunable diode-pumped Nd:YAG laser is used to generate the chirp signal. The system has been characterized over a measurement range from 1 to 11 m. In accordance with the state of the art of the FMCW interferometry, the average measurement accuracy of the system is 10-4 of the measured distance. The behavior of the frequency modulator has been experimentally characterized. A suitable signal elaboration procedure has been implemented to compensate for the nonlinear behavior of the optical frequency modulator.

In most length-measuring interferometers, the measuring retroreflector must move along a straight line. Only small deviations from this straight line are permitted; otherwise the interference signal becomes too weak. The interferometer can be made immune to movements perpendicular to the straight line by adding a plane mirror to it. The problem of alignment of the extra mirror, which is very critical in existing schemes, is solved by reflecting the reference beam in this plane mirror. However, the perpendicular movements partly alter the course of the optical path. The resulting loss in accuracy when using standard optical elements can be kept below 0.5 µm. With error compensation, the loss in accuracy can be kept below 0.1 µm. For interferometric measurement applications that do not require the highest attainable precision, this is acceptable. If the lateral position of the return beam is restored just prior to each measurement, the accuracy of this optical scheme is comparable to that of a standard one.

An extra-shortened He-Ne laser (shorter than 100 mm) does not work well because there is likely no longitudinal mode in the lasing bandwidth at some times and the output power rises and falls deeply from time to time. A method to solve this problem is presented and investigated. A piece of appropriately designed quartz crystal plate is used as a cavity mirror of a He-Ne laser, which has a larger longitudinal mode spacing than the lasing bandwidth to achieve an equal-spacing mode split. In other words, the number of longitudinal modes is doubled and the actual mode spacing is reduced to half of that without mode split. Therefore, there is always at least one mode operating in the lasing bandwidth and the power is stable at any room temperature. Such a laser with 85-mm cavity length and 0.24-mW output power is constructed.

We study numerically and experimentally various characteristics of a dye laser, using an optothermal bistable device as one of the cavity mirrors, when an etalon is inserted in the laser cavity. Firstly, we develop a model for describing this laser system. Secondly, we show that this laser can exhibit bistabilities in both intensity and wavelength simultaneously. Thirdly, we study one of the applications of this system: using it to improve the output power stability of a laser. Theoretical and experimental results show that, under proper conditions, the output power fluctuations at low frequencies which are produced by gain changes can be reduced remarkably by using this laser system. Finally, we study experimentally the switching dynamics in this laser, and demonstrate that optical flip-flop operation can be realized in it by positive pump light pulses, or by positive pump light pulses for onswitching and additional positive light pulses (to change the phase shift of the bistable mirror) for downswitching.

The morphology and the chemistry of dental enamel and dentin can be modified by irradiation with a CO2 laser to increase the acid resistance of the intrinsic mineral. The changes induced in hard dental tissues after laser irradiation are predominately determined by the photothermally induced temperature rise in the tissue. Therefore the temperature rise in the irradiated enamel and dentin must be determined under controlled laser conditions. Surface and subsurface temperatures were monitored after multiple-pulse CO2-laser irradiation at ?=9.3, 9.6, 10.3, and 10.6 µm with 1- to 20-J/cm2 pulses of 50- to 500-µs duration using radiometry and microthermocouples. Surface temperatures were significantly higher after 9.3- and 9.6-µm irradiation than for the more commonly utilized 10.6-mm CO2-laser wavelength. Permanent changes in the temperature response of enamel and dentin were observed at fluences greater than 2 J/cm2 and 100-µs duration for dentin and 5 J/cm2 for enamel. CO2-laser irradiation changes the thermal and the optical properties of these tissues, substantially changing the energy deposition for subsequent laser pulses. These changes affect both the amount of energy absorbed and the depth of absorption. The more efficient absorption at ?=9.3 and 9.6 µm may be advantageous for both cariespreventive treatments and ablation of exposed hard dental tissues while minimizing heat deposition in the tooth.

Light-induced scattering in a Ce:Fe:LiNbO3 crystal is highly temperature dependent. By considering the effect due to thermally activated ions in a photorefractive crystal, we have shown that the photoinduced space-charge field is neutralized by the activated ions. This neutralization in turn leads to a reduction of scattering noise in the crystal. By raising the crystal temperature, we have observed a reduction in scattering noise, which is consistent with our prediction. The signal-to-noise ratio (SNR) in a two-wave mixing amplification due to thermal activation effect is analyzed. Experimental demonstrations using a Ce:Fe:LiNbO3 crystal are provided in which we have shown that the SNR improves as the crystal temperature increases.

The day-to-day evaluation of a scanner system’s performance, a problem in quality assurance, is addressed by deriving a technique for establishing it’s system cutoff spatial frequency through the analysis of the edge slope. It is shown that the slope value at the midpoint of the system’s imaged edge (edge response) is related (and, for some conditions, equal to) its cutoff frequency. From the analysis, a parameter emerges that describes the sampling aperture as ‘‘hard’’ or ‘‘soft,’’ signifying whether its shape is best described by geometrical or physical optics, respectively. The physical implications of this description are discussed and results are substantiated experimentally using a microdensitometer.

A computer-based method for measuring the local refractive index gradient of phase objects using the projection moire´ technique is presented. A phase object is placed in front of the projected grating pattern. The image of the projected grating pattern deforms in accordance with the variation of the refractive index of the phase object. The moire´ fringe is generated by superimposing two images of the projected gratings pattern (the original pattern and the deformed pattern). The optical transmission characteristics of projection moire´ fringe is discussed. The moire´ deflectogram is analyzed by Fourier filtering and the phasemeasuring technique.

Deformable mirrors have been considered deficient for loworder- mode corrections such as tilt jitter. New materials and advanced computational procedures have made it possible to design a mirror to satisfy both stroke and bandwidth requirements for different aberration orders. Optimization techniques are applied to the systematic design of a multilayered deformable mirror composed of a bimorph polyvinylidene fluoride (PVDF) piezoelectric material. The deflection-bandwidth product is chosen as the figure of merit and maximized through the optimization process. Equations for the generated moment, stroke, frequency, and stress in a multilayered plate are identified. The limits on system performance are considered and treated as the design constraints. The design variables are chosen to be the geometrical dimensions of each layer and the applied electric field to the piezoelectric material. Discussion is presented on design, mathematical modeling, computer implementation, and application to a test model. The specifications are provided for a mirror with low-order wavefront correction capability. The successful procedure and results establish a basis for a general approach for the design of deformable mirrors of any shape.

A conception of an adaptive telescope optical system with the segmented primary spherical mirror having a diameter of 10 m is proposed. The system consists of four mirrors. The first two mirrors form a strictly afocal system with angular magnification of 4.2X. The second convex aspherical mirror has a diameter of 2.34 m. The next two mirrors, having diameters 2.52 and 0.51 m, respectively, form a focusing system with a focal length of 23.77 m. The focal length of entire system is 100 m; the length is 13.5 m, the angular field is 10 arcmin, its theoretical resolution is 0.1 arcsec. Alternatives for the main system arising due to changing of shape and relative position of mirrors of the focusing system are considered.

Formulas are derived for finite ray tracing through diffractive optical elements (DOEs). These formulas take a simple, Snell’s-law-like form, yet they are completely general, and can be applied to different types of diffractive lenses, including holographic optical elements and binary optical elements, which can be formed on surfaces of arbitrary shape. Additional formulas are developed for the calculation of wave aberrations introduced by the DOEs, and for pseudo-paraxial ray tracing through DOEs, which is used to determine the first-order properties of the system such as the Gaussian image position.

Holographic lenses in volume media suffer from the mismatch between the recording and the reconstruction wavelength. The lenses are recorded in the blue spectral range because of spectral sensitivity of the recording material. The lenses investigated here are reconstructed with red light. The aim of this work is the realization of diffraction-limited holographic lenses with a high diffraction efficiency in the presence of a wavelength mismatch. Therefore the lenses have to be recorded with at least one aspheric recording wavefront in order to fulfill the Bragg condition on the one hand and to correct the wave aberrations on the other. We propose an iterative optimization procedure which also eliminates the wave aberrations caused by the misalignments of the recording setup. This optimization procedure requires the accurate measurement of the wave aberrations. These aberrations are eliminated by an appropriate modification of the aspheric recording wavefront. In this paper the design principle, the recording setup, the interferometric measurement, and the iterative optimization procedure are given. This optimization procedure results in diffraction-limited holographic lenses.

A series of optoelectronic processing systems is under development at the Optoelectronic Computing System Center (OCSC) at the University of Colorado. The demonstrations consist of two facing optoelectronic modules that communicate with each other using bidirectional free-space optical channels. An analysis of the fabrication tolerances required for correct operation of these systems is presented. An overview of the system and the techniques used for fabrication is given, along with the tolerances achieved in practice. A comparison with theoretical results indicates the critical alignments within the system. Methods to obtain the alignments required for larger systems are discussed.

The motion adaptive spatial filter (MASF) is a kind of temporal filter proposed for noise reduction and temporal band limitation. The MASF uses motion vectors to extract temporal information in spatial domain. Therefore, inaccurate motion information causes some distortions in MASF operation. To decrease such distortions, a sophisticated motion estimation algorithm that can estimate motion with subpel resolution and afford to deal with sufficiently great motion is needed. Such algorithm, however, causes a large increment of the look-up table size required for the very large scale integration (VLSI) implementation of the MASF. To solve this problem, we analyze the effect of motion estimation error on the MASF and propose a postprocessing scheme for the estimated motion vectors. Experimental results show that by using the proposed scheme considerable amounts of distortions are eliminated without any increment of required look-up table size.

Substantial improvements of tunable diode laser absorption spectroscopy (TDLAS) with respect to detection speed and limits were obtained by introducing high-frequency modulation schemes, but the expected quantum-limited performance with optical multipass cells has not been attained yet on a routine basis. This paper is primarily devoted to the question of how refractive index fluctuations generated by a turbulent gas flow through an optical multipass cell affect the phase of a frequency-modulated laser beam and therefore influence the performance of highly sensitive spectroscopic measurements. It has been found that for measured pressure fluctuations of about 80 µbars in such a multipass cell, the expected sensitivity is limited to 7 X 10-7 in terms of optical density, which is more than one order of magnitude above the quantum limit. Further sensitivity improvement by signal averaging is limited by 1/f noise contributions from turbulence for integration times longer than 30 s. The consequences of the pressure fluctuations for absorption measurements with tunable diode laser-based systems are discussed with respect to state-of-the-art technology detection limits.

The effect of thermal defocusing of a pulsed Nd:YAG laser beam focused in a water cell is observed by a shutter video camera as a function of the position in the cell and the time during the laser pulse. In a first case the laser beam is directly focused into the water. In a second experiment the beam is coupled into a multimode fiber and the fiber output focused into the cell. A theoretical model is proposed explaining the observed phenomena. The simulation shows results that are in very good agreement with the measured values.

In its classical form particle image velocimetry (PIV) extracts two components of the flow velocity vector by measuring the displacement of tracer particles within a double-pulsed laser light sheet. The method described in this paper is based on the additional recording of a third exposure of the tracer particles in a parallel light sheet, which is slightly displaced with respect to the first one. The particle images resulting from these three exposures are stored on separate frames. The locations of the correlation peaks, as obtained by cross-correlation methods, are used to determine the projections of the velocity vectors onto the plane between the two light sheets. The amplitudes of these peaks are used to obtain information about the velocity component perpendicular to the light-sheet planes. The mathematical background of this method is described. Numerical simulations show the influence of the main parameters (e.g., light-sheet thickness, light-sheet displacement and out-of-plane flow component) on the resolution and reliability of the new method. A new recording procedure and its results will be shown to demonstrate the ease of operation when applying this technique to liquid flows.

A class of minimum duration-bandwidth product, discretetime, orthonormal wavelets is formulated using the discrete-time wavelet transform (DTWT) via the (L+1)-channel, perfect-reconstruction, discrete-time filter bank structure and a duration-bandwidth product measure averaged over the filterbank channels. We show a significant improvement in duration-bandwidth product over the discrete-time (filterbank) versions of Daubechies’s minimum-phase and least-asymmetric wavelets in our examples for filter lengths 8?N?22. These wavelets are expected to perform better than other classes of wavelets in applications where small duration-bandwidth product is important.

The discrete wavelet transform (DWT) is becoming increasingly popular as a powerful tool for signal decomposition techniques, which are important for analyzing nonstationary signals. Recently much emphasis has been placed on the time-frequency resolution properties of wavelets, which involves designing wavelets with good localization in both time and frequency via minimization of the time duration/frequency bandwidth product of the wavelet function. But certain practical applications, such as detection of edges in an image, require good time resolution only, which suggests minimizing only the time duration of the wavelet function. The results of minimization of the time duration of wavelets is described. As in Morris, Akunuri, and Xie (1995), the time duration of a wavelet function is expressed as a function of the wavelet-defining filter of length N. For a given N, the minimization is done over all possible orthonormal wavelets. The minimum duration wavelets (MDW) designed will be best suited for applications requiring good time resolution. Some examples along with comparisons with other wavelets are presented.

A wavelet preprocessed amplitude-modulated phase-only filter is implemented on both gray-scale and synthetic-aperture-radar (SAR) images. The algorithm, which exploits the multiresolution properties of a wavelet, is evaluated in terms of various pattern recognition statistics. The effect of thresholding nonlinearity is also studied.

Intensity modulation (IM), intensity-hue-saturation (IHS), and multiresolution wavelet decomposition (MWD) are ad hoc image merging techniques that enhance low-spatial-resolution color imagery with highresolution black and white imagery. While the MWD is a relatively new technique, both IM and IHS can be found in commercial software. These three merging techniques are compared to find which best retains spectral fidelity while adding spatial information. With this criterion, merging artifacts are identified and characterized. Specifically, the three techniques were tested with respect to resampling, misregistration, intensity variance, and complexity problems that arise in multiresolution image merging. Test images were used allowing the mean difference, standard deviation difference, entropy difference, and affected area to be calculated.

When digitizing an interferogram with a detector array, the sampling theorem requires the minimum local fringe spacing or period to be greater than twice the pixel separation. Thus, given a detector, there is a minimum fringe period that may be allowed.1,2 This minimum period, in turn, is set by the wavefront asphericity and the testing method.

When digitizing an interferogram with a detector array, the sampling theorem requires the minimum local fringe spacing or period to be greater than twice the pixel separation. Thus, given a detector, there is a minimum fringe period that may be allowed.1,2 This minimum period, in turn, is set by the wavefront asphericity and the testing method.

The book by Nils Carlson is a comprehensive, if somewhat uneven, treatment of the efforts over the last decade toward realizing monolithic diode-laser arrays of high coherent power (>1 W).

This PDF contains the section with book reviews including: "Introduction to Semiconductor Integrated Optics. By Hans P. Zappe"