Almost two years to the day after the first meeting on OSA-SPIE unification took place, the proposal was rejected by OSA members and narrowly approved by SPIE members. Both memberships returned nearly identical results, with 51% of the voting SPIE members and 49% of the voting OSA members supporting the merger. Although it passed on the SPIE vote, though just barely, the merger will not take place because of the negative OSA vote.

Players of a video game may sometimes find the use of conventional interfaces inappropriate. In such cases, we think that interfaces realized with a vision-based gesture recognition system may find favor. The artificial retina (AR) chip is a versatile image sensor whose use ranges from normal image acquisition to on-chip image processing, including on-chip image convolution. In this paper, we describe a gesture- input video game system, with the AR module including the AR chip, and motion-based gesture recognition algorithms. We showed that the algorithms can be accelerated by projection data, the direct output from the AR chip. To show its performance, we have applied our system to two commercially available video games.

In micro-domain applications, such as assembly of hybrid microelectromechanical systems, optical microscopes form a critical component of visually based microassembly systems. In order to effectively use vision feedback to guide manipulation strategies using a broad range of optical microscope configurations, accurately calibrated parametric models of optical microscopes are required. Although many calibration techniques exist for macroscopic camera-lens systems, there is a dearth of literature on calibration of parametric microscope models. Optical microscope calibration has unique characteristics that are quite different from normal camera calibration, including (1) unique calibration parameters; (2) calibration patterns that must be parallel to the image plane, and (3) restriction to single-plane calibration. In this paper, we propose a parametric microscope model and calibration algorithm specifically for optical microscopes. Our approach extends existing camera calibration techniques to include the unique parameters of optical microscopes and also allows the use of a single calibration plane perpendicular to the optical axis. The validity and accuracy of our proposed microscope calibration method are tested by experiments.

We analyze and discuss the error generated by an algorithm commonly used in computer color matching and point out the necessity of the calculation used for recipe revision. On the basis of this calculation a model for computer color matching used in recipe revision is proposed that will cause the computer color-matching technology to be used more widely and to meet the demands in practice.

The full potential of second-order nonlinear polymers can be utilized in electro-optic polymer modulators with a DC biased operation scheme to greatly reduce the half-wave voltage. This technique makes use of the total achievable electro-optic coefficient, which can be more than three times the value that is used by the conventional devices of poled electro-optic polymer. As the result of the DC bias and with high-?? chromophores, a low half-wave voltage of 1.5 V was achieved with 2-cm-long birefringent waveguide modulators at the wavelength of 1.3??. Results of a 200°C stability experiment indicate that this scheme also enables electro-optic polymer devices to meet the short-term high- temperature stability requirement because the polymer does not need to be poled prior to high-temperature steps.

Hollow glass waveguides provide an attractive means for the delivery of IR laser power and for broadband temperature and chemical fiber sensor applications. These waveguides are made using liquid- phase chemistry techniques to deposit a thin dielectric film of Agl over metallic Ag on the inside of small-bore silica tubing. By controlling the thickness of the Agl film, we can tailor the optical response of the guides to give low loss at selected IR laser wavelengths or to produce a broadband response. In either case, the losses depend on the thickness uniformity of the Agl film. Initially, the Agl film thickness varies as much as 14% from end to end of a 6-m long guide. A faster Agl deposition rate reduces this variation to only 1.3% over a 5-m long waveguide. This leads to a reduction in the total loss for the hollow guides in the 2 to 12 ?m region.

A new image-processing approach for object analysis in life and earth sciences is presented. This approach is based on a multiresolution algorithm in image processing. A clamshell surface has been digitized using a noncontact optical sensor based on laser triangulation. The 3-D surface obtained constitutes an image that can be characterized by multiresolution analysis. The application of this method to the study of a bivalve shell surface (Unio sp., Recent Atlantic, Holocene) allowed the various growth increments and their potential relationship with environmental constraints to be measured. The algorithm used in this paper is based on the wavelet transform theory.

We propose an approach to realize distortion-invariant pattern recognition for segmented 2-D patterns with an axis of symmetry. Parameters of distorted input pattern are first estimated and normalized. Invariant pattern recognition is then achieved by correlating the normalized input pattern with the undistorted reference, using a classical matched filter or its modified version. This approach realizes pattern recognition invariant to rotation, scale, and projection without the need to synthesize a distortion-invariant filter. Simulation results demonstrate the effectiveness of this approach.

We present an object-based fractal coding scheme based on the partitioned iterative function system (PIFS) that is similar to the standard fractal image coding schemes. The partitioned blocks in our object- based fractal coding scheme are classified as boundary blocks and internal blocks according to their shapes. The boundary blocks are located in the boundary region of an object and their shapes are arbitrary, whereas the internal blocks are located in the internal region of an object and their shapes are square. As the internal blocks do not affect the blocks to appear in the process of the ordinary fractal coding except that the domain pool is not an image but an object, encoding for the internal blocks follows the ordinary fractal coding schemes as they are. However, the boundary blocks should be carefully handled, as they have properties that make them different from the internal blocks, i.e., they have arbitrary shapes and their domain pool is composed of the internal blocks not the boundary blocks. Considering these properties, we propose three methods to encode the boundary blocks by making an alteration to the shapes of range blocks and domain blocks, and a method to encode them using the predecoded internal blocks. The methods are mathematically analyzed and their performance is experimentally evaluated.

We propose a Fourier-based interpolation method designed to suppress the oscillatory artifacts that occur when interpolating an undersampled function. In the proposed method, we formulate the problem as an optimal recovery problem incorporating a priori knowledge of the aliasing. Our computer simulations demonstrate that the proposed method can produce good interpolation results even when the functions are severely undersampled.

We develop a laser-diode phase-shifting interferometer that can capture frame-rate phase-shifted interferograms on a video tape with a high capacity for their storage. The number of the video frame is simultaneously recorded on an audio track of the video tape. In a four- step manner, the injection-current variation of the laser diode is synchronized with the field pulse of a CCD camera to produce ?/2-step phase shifts at a frame rate of 1/30 s. An intensity distribution of the interfero- gram on any frame can be determined through a video tape with a frame memory using the recorded frame number. The four successive phase- shifted interferograms are used to calculate a distribution of frozen phase by a phase-extraction algorithm. The phase error due to the frame-byframe intensity repeatability is theoretically analyzed.

Noise reduction and improved resolution are some serious challenges for precision measurement in the era of nanotechnology. Although several factors that lead to the loss of accuracy in precision measurement system were realized and some were eventually eliminated, some critical factors that were overlooked remain. One such factor, which has become an inherent part of the system, is that caused by the vibration of the acousto optic modulator (AOM) in heterodyne interferometry. We highlight the effect of error induced by the vibration of the AOM and, different methods of eliminating this error by the modification of the optical layout discussed.

We describe the principle of a motionless interferometer based on the Talbot effect. The test object is scanned by a convergent spherical beam with variable curvatures and lateral shifts. The wavefronts are generated by a programmable off-axis Fresnel zone plate, implemented on a high resolution and high speed binary phase spatial light modulator. The aberrated field is then analyzed by passage through a fixed binary phase grating, detection by a sensor located in a fractional Talbot plane, and numerical phase retrieval from the Fourier analysis of the reconstructed intensity distribution. We describe the principle of the object multiresolution analysis. The system operation is then illustrated experimentally in a simple case exhibiting subwavelength resolution capability. Finally, we give some orders of magnitude for the dynamic range that can be expected with such a system, as well as possible extension toward very high resolution measurements.

A rectangular flat metal plate is made to randomly vibrate when excited by a shaker. A digital holographic interferometry out-ofplane optical system is used to investigate the temporal development of the induced transient deformations. The system relies on two and four laser pulses emitted within a single envelope of the flashlamp contained in the oscillator cavity of a ruby laser, and also on three CCD cameras used to record two or four digital images resulting from the same number of holographic interferograms. Hologram data reconstruction is performed digitally and Fourier evaluated to obtain quantitative deformation data for each of the formed interferograms. The deformation data is broken down into resonant mode components, sums of which are used to computer reconstruct the experimental deformation data. The error produced when using a finite sum of resonant mode components is considered. Experimental and computer reconstruction results for the transient vibration deformation of the plate are presented and compared.

We describe two new techniques for spatial phase stepping using a computer-generated holographic optical element (HOE). The techniques are implemented with electronic speckle pattern interferometry, although they could be applied to other optical metrology methods. The first technique uses an HOe that introduces a known phase step between the ±1 diffracted orders, without being translated. The interference phase corresponding to object deformation can be calculated from a single TV frame, and the technique is therefore suitable for the measurement of transient deformations. We apply the technique to measure the phase changes caused by refractive index variations in an evolving thermal plume. The second technique requires the HOE to be used in conjunction with a phase grating. In this technique, the HOE and the phase grating are used sequentially for the two exposures required in speckle interferometry (made before and after object deformation) and is therefore suitable for making dynamic deformation measurements with a double-pulsed interferometer. A root mean square (rms) precision of 2?/20 rad for the phase measurements is obtained.

Phase unwrapping by a lookup table method has been used to analyze the phase map with singular points that is derived from a multiple-spatial-carrier-frequency fringe pattern. The phase unwrapping can be performed independently on a pixel-by-pixel basis without reference to the phase values of neighboring pixels, and is relatively insensitive to errors. Experimental results are presented that demonstrate the validity of the principle.

An enhancement to a previously developed Karhunen-Loeve/ discrete cosine transform-based multispectral bandwidth compression technique (Saghri et al., 1995) is presented. This enhancement is achieved via the addition of a spectral screening module prior to the spectral decorrelation process. The objective of the spectral screening module is to identify a set of unique spectral signatures in a block of multispectral data to be used in the subsequent spectral decorrelation module. The number of unique signatures found depends on the desired spectral angle separation, irrespective of their frequency of occurrence. This set of unique spectral signatures, instead of the signature of each and every point in the block of data, is used to construct the spectral covariance matrix and the resulting Karhunen-Loeve spectral transformation matrix that is used to spectrally decorrelate the multispectral images. The significance of this modification is that the covariance matrix so constructed is not entirely based on the statistical significance of the individual spectra in the block but rather on the uniqueness of the individual spectra. Without this added spectral screening feature, small objects and ground features would likely be manifested in the low eigen- planes mixed with all of the noise present in the scene. Since these lower eigenplanes are coded via the subsequent Joint Photographic Experts Group (JPEG) compression module at a much lower bit rate, the fidelity of these small objects is severely impacted by compression- induced error. However, the addition of the proposed spectral screening module relegates these small objects into the higher eigenplanes and hence greatly enhances the preservation of their fidelities in the compression process. This modification alleviates the need to update the covariance matrix frequently over small subblocks, resulting in a reduced overhead bit requirement and a much simpler implementation task.

We describe and present experimental results of a nonconventional process that can be used for the polishing of optical surfaces. In particular, we present data regarding polishing uniformity, removal rate, surface microroughness, and subsurface damage.

A truly portable scanner would add another dimension to mobile computing. Using the compact optical configuration single optical fiber imaging (SOFI), a handheld scanner can be built inside a penshaped housing. Here, an image sensor formed on a glass substrate is used. Each photodiode of the image sensor has multiple apertures coupled to multiple optical fibers. The light is transmitted through the glass, the apertures and the fibers successively, and a document is illuminated. Scattered light comes back in the same fibers and is detected by a sensitive part of the photodiode. Therefore, a single fiber is involved to both illuminate the document and detect the reflected light. By incorporating light-absorbing material at the fiber boundaries, optical crosstalk inside the fiber array is suppressed. The design of this SOFI-based scanner is studied in detail. A prototype is developed: the width of its housing and the input length are 13 and 110 mm, respectively. The key components are an image sensor fabricated by a-Si and poly-Si thin film technology, a 2 mm-thick array of optical fibers with 0.9 numerical aperture, and an array of three-color LEDs. Images of color documents and fingerprints are successfully acquired by this scanner.

A novel optical technique has been developed to measure particle diameters in multiphase flows with high accuracy. This technique uses a photodiode array to measure the spatial frequency of the laser light scattering from a spherical particle located in the measurement volume. An optimized optical orientation angle is introduced by taking two scattering mechanisms into consideration, which provides high accuracy and avoids the measurement-volume effect in sizing large particles. The performance of the new method was simulated by using the generalized Lorenz-Mie theory (GLMT) and validated by experiments. The results show only 2% measurement uncertainty for particle sizes as large as 1200 ?m, which are commonly encountered in gas-solid and bubblywater multiphase flows. The feasibility of this technique was demonstrated experimentally.

Human face location is an important basis for automatic face recognition systems. We define one kind of symmetry transform called the directional symmetry transform (DST) that can provide more information for shape analysis in addition to the characteristics of symmetry. When applied to object detection, it can provide much more convenience. One fast algorithm based on the DST is proposed for face location from frontal face images with simple backgrounds. The experimental results show its effectiveness.

we observe that conventional electrically addressable spatial light modulators have different transfer functions along the fast (horizontal) and the slow (vertical) directions. We then propose to use the directionality of a spatial light modulator to increase the performance of a joint transform correlator. Our experimental results show that input space- bandwidth product of a joint transform correlator can be significantly increased by recording a hologram so that interference fringes run along the fast (horizontal) direction of a spatial light modulator.

We present a comprehensive theoretical model suitable for treating the effect of pulsed collinear photothermal deflection spectroscopy (PDS). The work is an extension of the theoretical model previously developed for the mirage effect, which can take into account both photothermal deflection and photothermal diffraction effects based on the Fresnel diffraction theory. With the diffraction model, both the collinear PDS and the photothermal lensing spectroscopy techniques can be treated in a unified manner. The model provides a detailed analysis of the laser-induced optical diffraction effect and can be used to optimize experimental parameters. The modeled results are presented in detail, with an emphasis on the advantages of using a near-field detection scheme for achieving the best sensitivity to local temperature change and better experimental stability against environmental noise.

Good quality AgGaSe2 single crystal 22 mm in diameter and 80 mm in length is grown by an improved technique, i.e., descending ampoule with rotation. Although the crystals grow with optical defects, postgrowth heat treatment procedures are used to successfully eliminate the defects and produce material of 62.4% IR transmission at 10.6?m. The frequency-doubling device is cut from a cylindrical crystal boule using a new orientation cut method according to the type-1 phase-matched condition. A high optical quality, orientated AgGaSe2 single crystal of 10x 10x 12 mm size is used for second harmonic generation (SHG) of the pulsed output of a 10.6?m CO2 miniature transversely excited atmospheric laser. A relatively high energy conversion efficiency of 12% is obtained. It is found that the optical surface damage threshold of the crystal is larger than 12 MW/cm2.

We examine metrics for measuring clutter effectiveness on model-based automatic target recognition (ATR) systems with forwardlooking infrared (FLIR) sensors. The measure for clutter effectiveness proposed is the difference of two Kullback-Leibler distances between the idealized approximate probabilistic models without clutter and the real model containing clutter. We establish that occluding objects and clutter, when manipulated, do not present a fundamental challenge to model- based ATR systems if the model manipulated is an accurate representation of the obscuring clutter. However, if the obscurer is not manipulated, performance degrades in cases where the obscurer is an ''effective clutterer. To quantify the effect of clutter in ATR, estimation and detection problems are considered for rigid ground-based targets. For estimating the orientation of a vehicle, the Hilbert-Schmidt distance is employed.

we propose an automatic target recognition (ATR) algorithm that uses a set of dedicated vector quantizers (VQs) and multilayer per- ceptrons (MLPs). For each target class at a specific range of aspects, the background pixels of an input image are first removed. The extracted target area is then subdivided into several subimages. A dedicated VQ codebook is constructed for each of the resulting subimages. Using the K-means algorithm, each VQ codebook learns a set of patterns representing the local features of a particular target for a specific range of aspects. The resulting codebooks are further trained by a modified learning vector quantization (LVQ) algorithm, which enhances the discriminatory power of the codebooks. Each final codebook is expected to give the lowest mean squared error (MSE) for its correct target class and range of aspects. These MSEs are then input to an array of window-level MLPs (WMLPs), where each WMLP is specialized in recognizing its intended target class for a specific range of aspects. The outputs of these WMLPs are manipulated and passed to a target-level MLP, which produces the final recognition results. We trained and tested the proposed ATR algorithm on large and realistic data sets and obtained impressive results using the wavelet-based adaptive product VQs configuration.

Three-dimensional (3D) digital images arise in many radiologic imaging applications. In an application such as virtual endoscopy, interactive navigation through the 3d image data—i.e., the interactive creation of an animated sequence of views—can reveal more information than single static two-dimensional (2D) views. For this case, the 3D image acts as a "virtual environment," representing the imaged structures, and a computer-based system can serve as a tool for navigating through the environment. But such navigation must occur at interactive speeds if it is to be practical. This demands fast volume visualization from arbitrary viewpoints. We present an inexpensive, fast, volumerendering method that can generate a sequence of views at interactive speeds. The method is motivated by a template-based approach and the temporal-coherence concept. The method forms part of an integrated system for virtual-endoscopic exploration of 3D radiologic images. Pictorial and numerical results illustrate the effectiveness of the method.

Optical image recognition of three-dimensional (3-D) objects has recently been proposed. The proposed system basically performs correlation of the holographic information of a 3-D reference object and that of a 3-D target object, and hence 3-D matching is possible. However, the system is not 3-D shift-invariant. A solution for 3-D shift invariance has been pointed out, but has the drawback that only target objects that are farther away than the reference object can be shift-invariantly detected. We propose a method by which 3-D shift invariance is possible without that drawback. The method applies Wigner analysis to the power-fringe-adjusted-filtered correlation output to extract the 3-D location of the matched target object. We analyze the proposed method and present computer simulation results.