This PDF file contains the editorial “Tower of Google II” for OE Vol. 45 Issue 03

Image mosaicking is a procedure of integrating information from a series of images to create a comprehensive view of the scene. It is typically carried out by selecting a subset of pixels from each of the individual images, matching these selected pixels from different images, and then mapping all the images onto a common image grid. The number of selected pixels is a critical parameter that affects both computational complexity and mosaicking accuracy. An image mosaicking algorithm is developed by using a novel dynamic point selection concept. The algorithm automatically determines the number of pixels to select according to the similarity of the images. Simulations show that the proposed algorithm generates mosaic accurately and efficiently.

Compact aplanatic concentrators can reconstitute the flux of a near-field ultrabright incoherent light source and couple it into an optical fiber. Performance near the thermodynamic limit for light transfer can be realized, even at high numerical aperture. The prospect of arc-discharge lamps as alternatives to lasers for many surgical fiber-optic procedures motivate this investigation. LED-fiber and fiber-fiber coupling constitute additional applications. The contours of these achromatic mirrored systems are analytic functions, which facilitates rapid surveying of a wide range of design options.

For fabricating the metallic mesh on a spherical substrate, a method that can adjust the center of a sphere to its spin axis is presented. A grating is used for the light splitting and uniting, and fixed on the center of spherical substrate. First, a laser beam is diffracted by the grating and splits several beams, and a ±1-order beam of diffracted beams arrives the substrate and returns to the grating. They are diffracted by the grating again, and an interference pattern is obtained in the overlapped region of two diffracted beams. A polarization prism and a 1/4 wave plate deflect the coherent beam from former incidence direction, and the interference pattern is displayed on a screen by the transform of a CCD detector. While the curvature center of the spherical substrate aligns with its spin axis, the interference pattern is still when the substrate is revolving about its spin axis, or else the interference pattern will alter with the revolving of substrate. The change scale of interference fringes also corresponds with the error of alignment.

We present a perceptually-adaptive in-band preprocessing scheme for 3-D wavelet video coding. In our scheme, after the original video is decomposed by 2-D spatial wavelet transform, a preprocessor is incorporated to remove some visually insignificant wavelet coefficients (noise-like) before the motion compensated temporal filtering of each spatial subband. The preprocessing process is guided by a wavelet domain just-noticeable-distortion profile, which locally adapts to spatial wavelet transform coefficients. Experimental results show that the proposed scheme can efficiently enhance the visual quality of coded video with the same objective quality at different bitrates.

We introduce an intelligent optical cross connect (OXC) structure based on our lambda-group model, as well as a dynamic algorithm of configuration helpful to complicated optical traffic grooming, which presents a distinctive approach of dividing granularities into specific tunnels for effective treatment. Results and figures from experiments show that the particular channel partition not only is helpful to port reduction significantly, but also improves the channel and blocking performance for dynamic connection requests.

Carbon fiber composite technology for lightweight mirrors is gaining increasing interest in the space- and ground-based astronomical communities for its low weight, ease of manufacturing, excellent thermal qualities, and robustness. We present here first results of a project to design and produce a 27-cm diameter deformable carbon fiber composite mirror. The aim was to produce high surface form accuracy as well as low surface roughness. As part of this program, a passive mirror was developed to investigate stability and coating issues. Results from the manufacturing and polishing process are reported here. We also present results of a mechanical and thermal finite element analysis, as well as early experimental findings on the deformable mirror. Possible applications and future work are discussed.

This paper describes a method and system for the measuring the two-dimensional distribution of birefringence dispersion. An optical arrangement consists of a white light source, parallel polarizers, a CCD camera, and an acousto-optic tunable filter for selecting wavelength of the incident light. The intensity of spectroscopic polarized light changes sinusoidally as a function of wave number, and its period changes slightly because of birefringence dispersion. The fast Fourier transform method is used to analyze the birefringence dispersion from the spectroscopic polarized light. One hundred twenty-eight captured images are used for the analysis. Some experimental results on 2-D birefringence dispersion distributions are shown for the demonstration of this method.

A measurement system utilizing laser frequency splitting is presented for the precise measurement of retardations of arbitrary wave plates. The wave plate to be measured is inserted into a laser resonance cavity, and its retardation can be determined by measuring the difference of split frequencies and longitudinal mode spacing of the laser. In the system, mode spacing measurement is performed under multimode oscillation according to the relation between polarization directions of modes. To solve the problem of strong mode competition in measuring half-wave plates, a transverse magnetic field is applied to weaken the competition so as to make the system applicable even when the frequency difference is less than the threshold for strong mode competition. The error of the system is less than 5 arcmin, and standard deviations of about 1 arcmin are obtained from repeated experiments samples of quarter- and half-wave plates. No element rotation is required during measurement, and no mechanical quantity, such as angle, needs to be measured.

The self-calibration test known as the random ball test (RBT) is adapted and applied to instrument calibration for measurements of microrefractive lens figure error. The RBT exploits the symmetry properties of a microsphere, resulting in a low-uncertainty estimate of the instrument biases. One hundred surface patches on a 1-mm-diam steel sphere are imaged by commercial instruments then averaged together in software to determine the instrument bias for a 500-µm radius of curvature test piece. The results show biases on the order of a few hundred nanometers peak-to-valley for a scanning white light interferometer and a Twyman-Green interferometer.

We investigate the real-time velocity profile for optical tomography using the sensor-to-sensor (STS) and pixel-to-pixel (PTP) methods. Although both methods are similar, the pixel-to-pixel velocity is more accurate than sensor-to-sensor velocity because it measures the pixel velocity within all areas of the pipe and the input signals are from upstream and downstream concentration profiles. The sensor-to-sensor velocity measures only the peripheral velocity of pipe. This velocity method obtains the input signals from the receiver output voltage of upstream and downstream sensors. Problems occur during the nonflow condition because the receiver may have noise, resulting in an incorrect velocity, which is not 0 ms^–1. This kind of problem does not happen in the pixel-to-pixel velocity method because all of the noise obtained is filtered by the image reconstruction algorithm used in this project. The velocity profile can easily be created by the pixel-to-pixel velocity method.

We conduct displacement/strain measurements on the microscale using light microscopy and digital image correlation (DIC). Errors in the measurements attributed to the optical arrangement and aberration induced at high magnification are identified using a warping function. Coefficients of the warping function are determined using a simple technique that employs a precisely made orthogonal cross-grating plate. By acquiring images of the grating and identifying the nodes using subpixel techniques, a relationship between the object and the image planes is established. Thus, the displacement/strain derived by means of DIC is corrected by converting the displacement components in the image plane to the coordinate system existing on the object's surface. The approach is validated through a determination of the elastic properties of common metals; errors in estimation of the elastic modulus were within 4%. Although surface preparation generally plays a critical role in successful application of DIC, it is found to be of minimal importance under high magnification. Instead, the natural surface texture can be used with adjustment of the light incident angle. Results of the study show that DIC is a powerful tool in performing displacement/strain measurements on the microscale using a light microscope provided that an adequate correction is employed for image distortion.

An improved method is proposed to perform contouring on column basing in phase-measuring profilometry. The four images including body, body and grating, grating, and the reference plane are acquired by a digital camera. The shade area and the shape area of body are also obtained with the processing between four images. The main axis of the projector and the main axis of camera are not crossed and are also not in the same plane to arrange the measuring system easily and conveniently. The unwrapped phase is also easily obtained with the help of an adaptive bandpass filter and a modulation function because phases between adjacent unwrapping points are continual (in generally <2π) in the body area of image. The formulas for the phase subtraction and the geometric points are first provided for the no-cross-axis technique. The experimental results show that this technique is available for practical applications.

A systematic investigation of Ag_xSb_2(1–x)Te_3(1–x) (x=0.16, 0.18, and 0.20) is reported. The optical properties of the material were studied using a UV spectrophotometer. The optical behavior of a material is generally utilized to determine its optical constants such as reflectance, transmittance, and optical band gap E₀. The value of E₀ is estimated for these samples by measuring the absorption coefficient as a function of the wavelength of light in the range 300 to 1400 nm. Films of different Ag content were studied both before and after annealing. It was concluded that this material is a candidate for phase change optical memory.

The modulation transfer function (MTF) of a pixelated digital micromirror device spatial light modulator has been defined and derived as a function of the fill factor and the shape of the micromirror. Based on the formula, we have investigated the dependence of the MTF on the orientations and spatial frequency of the input patterns.

LED-based projectors have numerous advantages compared to traditional projectors, such as compactness, larger color gamut, longer lifetime, and lower supply voltage. As LEDs can switch rapidly, there is the possibility to pulse. However, there is also an important disadvantage. The optical power per unit of étendue of an LED is significantly lower than, e.g., an ultra-high-performance (UHP) lamp. This problem can be remedied partly by pulsing the LEDs. If one drives an LED with a pulsed current source, the peak luminance can be higher, albeit the average luminance will not increase. By pulsing two LEDs alternately (50% duty cycle), their increased flux can be added up in time and will generate a higher average flux within the same étendue. We combine the LEDs with a polarizing beam splitter (PBS) and change the polarization of one LED with a switchable retarder. The achieved substantial net gain after all losses is 36%.

The development of a photovoltaically (PV) powered laser communication system that constitutes a miniature, highly energy-efficient wireless communication technology is described. The technology is based on the direct integration of a multiquantum well (MQW) modulating retroreflector (MRR) optical communication node and a monolithically integrated module (MIM) PV power source. The MQW MRR optical data link exploits the shift in the MQW absorption peak under an applied reverse bias to modulate incident laser light, enabling binary encoding of data for transfer. A MIM consists of many individual solar cells monolithically integrated on a single substrate and offers the design versatility necessary to enable efficient electrical conversion of both incident sunlight and the system laser light and the ability to match the voltage output to the MRR requirements. A description of the development of the MRR and MIM components of the system is given. Results of bench-top demonstrations of the operational system are presented.

To analyze mode fields in a seven-folded stable resonator with circular diaphragms, we introduce a new method combining the traditional self-reproducing theory with diffraction theory in the form of the ray matrix. Since the iterative method presented by Fox and Li is complex for the seven-folded resonator with diaphragms, in this paper, by means of Collins's formula in polar coordinates, diffraction integral equations expressed by ray matrices are converted to finite-sum matrix equations along diffraction interfaces. Moreover, diaphragms and reflective areas in the seven-folded resonator are simplified to apertures given by the pupil function. Finally, using the self-reproducing principle, we describe mode fields and their losses in the folded resonator as eigenvectors and eigenvalues of a transfer matrix. By calculating the eigenvectors and eigenvalues of the transfer matrix, we obtain eigenmode field distributions and their losses in the resonator. It is shown from simulation results that the seven-folded resonator with circular diaphragms can easily yield the fundamental mode, so it can output laser beams of good quality.

We demonstrate the feasibility of a wavelength-switchable fiber-Bragg-grating external-cavity semiconductor laser source for direct modulation with active wavelength stabilization. The laser device works with two selectable wavelengths, can be directly modulated at 2.5 Gbit/s, and shows wavelength stability down to ±14 pm.

A description of a novel Y-branch photonic crystal laser diode operating at 1.55 µm is presented. Optical gain calculations within an InGaAsP-InP multiple-quantum-well (MQW) gain medium predict a threshold current of 0.5 mA, which is lower than a typical ridge laser. The resonant structure of the beamsplitting laser is modeled using a scattering matrix approach incorporating the effects of multiple internal reflections on the cavity mode structure and mode spacing. Phase-matching conditions are derived, taking into account all the different resonant substructures. The results from the phase-matching analysis show an operating mode that reduces threshold current beyond that of a geometrically similar conventional ridge laser. This Y-branch photonic crystal laser, when cascaded as a multichannel resonant structure, produces numerous optical intensity levels that can be used in an optical analog-to-digital converter (ADC).

An advanced design of an all-fiber passively mode-locked fiber ring laser is reported. By combining a fiber Bragg grating loop mirror with a segment of unpumped Er/Yb-codoped fiber as a saturable absorber, a stable continuous-wave mode-locking pulse with an even amplitude and a lowered noise background is obtained. The pulse width is narrower than 4 ns, with a repetition rate of 7.94 MHz. A high average output power of 319 mW and a slope efficiency of ~11.7% are achieved.

A high-power AlGaInP laser diode with current-injection-free region near the facet is successfully fabricated by metaorganic chemical vapor deposition (MOCVD) using the (100) direction n-GaAs substrates with a misorientation of 15 deg toward the (011) direction. The maximum continuous wave output power is about 90 mW for the traditional structure. In comparison, the maximum output power is enhanced by about 67%, and achieves 150 mW for LDs with current-infection-free regions. The fundamental transverse-mode operation is obtained up to 70 mW. Output characteristics at high temperatures are also improved greatly for an LD with a current-injection-free region, and the highest operation temperature is 70 °C at 50 mW without kink. The threshold current is about 33 mA, the operation current and the slope efficiency at 100 mW are 120 mA and 0.9 mW/mA, respectively. The lasing wavelength is 658.4 nm at room-temperature 50 mW.

An approach to long-range 3-D imaging using laser illuminated range-gated viewing is presented. The basis for 3-D scene reconstruction is an image sequence acquired using a sliding gate delay time. Two different methods are suggested, and algorithm performance is investigated through Monte Carlo simulations using a simplified system and imaging model. Assumptions are justified by comparison with real measurements at range of 0.8 to 7.2 km. It is shown that range resolution and precision become significantly better than system design parameters such as gate length, gate transition length, and gate step length. The presented reconstruction methods thus enable high-precision range imaging using available long-range gated imaging systems.

We describe new light-driven actuators based on films of polymer polyvinylidene fluoride, known as PVDF. The actuators employ the photomechanic bending of the polymer film caused by low-power (10 mW and less) laser radiation. The photomechanic effect combines various physical mechanisms, such as thermal expansion, the converse piezoelectric mechanism, and the photovoltaic and pyroelectric mechanisms, while the mechanism of thermal expansion is dominant. The force applied by the actuators to external objects is measured with a torsion balance. It is proportional to the power of laser beam and could be as high as 10^–4 N for a 50-µm film illuminated with a 10-mW laser beam. We demonstrate mechanical vibrations of a 1×7-mm strip actuator at a frequency of 0.3 kHz. As examples of possible applications, a photonic switch and an actuator with a closed-loop motion that could drive the inner workings of a conventional mechanical clock were demonstrated. The proposed actuators have a potential of being used as propulsion components of future light-driven micro/nano systems.

We present a dynamic bandwidth allocation scheme for Ethernet passive optical networks. This scheme, referred to as TDBA, is based on efficient threshold reporting and provides quality of service (QoS) to three different classes of packets. It is shown that network resources are efficiently allocated among the three traffic classes while guaranteeing the requested QoS by avoiding nearly all fragmentation losses. Simulation results using OPNET show that the TDBA scheme provides better packet delay and channel utilization than the other previously published allocation schemes.

A highly flexible optical packet compressor is presented. The compressor is capable of providing various compression ratios using the same hardware. It is composed of three parts: a chirped packet generator, a signal compressor, and a wavelength converter. In the chirped packet generator, the modulator intensity-modulates a series of supercontinuum chirped optical carriers and generates a series of chirped optical packets. The signal compressor compresses the chirped packets. The wavelength converter then transforms the compressed wideband optical packets into single-wavelength signals. We numerically demonstrate that using dispersive devices (chirped Bragg grating array or dispersion compensation fiber), we can compress both the width of pulses and the distance among pulses at the same time. This results in an increase of the bit rate. We also show that during the compression, the optical packet suffers distortion in the time domain, which can be defined as an extinction ratio. The distortion can be minimized by control parameters such as carrier chirping and modulating bandwidth. We present a highly flexible optical packet compressor, which is capable of compressing hundreds of bits packets from low speed (mega- or gigabits per second) to very high speed (up to 40 Gbits/s).

Multiple-beam Fizeau fringes are applied to study the effect of pure mechanical bending on some of single-mode optical fiber parameters. The induced birefringence and the associated radial refractive index profiles are measured as a function of fringe shift in case of very sharp radii of curvature. For the first time, the measured variation in index profiles incorporation with the minimum variance technique are employed to study some important parameters such as the oscillation energy, dispersion energy, and lattice energy profiles of the fiber cladding material predicting their radial variation with bending. The material dispersion profiles are specified as a function of the evaluated variations of the radial profiles of the oscillation and dispersion energies, respectively. The presented method provides an accurate study of the fiber parameters and their variations due to external effects, avoiding at the same time the complications of the dispersion measurements. Microinterferograms are given for demonstration.

A wavelength-sensitive variable fiber optic attenuator (VFOA) is introduced by incorporating the movable optical mirror with a fixed desired thin film filter (TF) in a compact in-line reflective architecture. Our TF-based wavelength-sensitive VFOA features low component count, high attenuation dynamic range, and high optical isolation. Experimental proof of concept using an off-the-shelf measured 1545.71-nm TF and a mechanically movable mirror shows a 4.64-dB optical loss, and a low polarization-dependent loss of 0.12 dB at the output port. A high optical attenuation dynamic range of >43 dB with a better than 30-dB optical isolation is obtained. Several proposed reflective wavelength-sensitive TF-based VFOAs can also be cascaded to form a low-cost, compact, multiwavelength amplitude controller.

We present new approaches for realizing enhanced tunable filters based on filling special fibers with nematic or cholesteric liquid crystal. One of the suggested approaches is to realize high-resolution longitudinal structure of electrodes using printed circuit board technology and then fill tapered capillaries with the liquid crystal, thus obtaining the desired spectral response at low voltage and with low insertion loss. Another approach is to fill special photonic crystal fibers with liquid crystal and obtain spectral tunability by generating phase differences between light paths.

A two-step computational technique for object reconstruction from digitized in-line Fresnel holograms of particle fields is proposed that eliminates the preliminary stage of depth extraction, when the goal is recovery of only the object cross-sectional shape. The first step involves a space/spatial-frequency domain chirp separation algorithm based on a two-dimensional (2-D) separable space-varying filtering scheme to separate a hologram into its single-object components. As the second step, a recently proposed Fourier synthesis technique is applied to separated components to reveal the cross-sectional morphology of individual 2-D objects, if they are sufficiently small as compared to the hologram size. The proposed method and a modified version of it based on 1-D filtering for performance improvement are illustrated on synthesized holograms. These simulations indicate that closely spaced planar objects can be recovered without significant dispersion, and back-to-back objects can also be reconstructed with tolerable leakage if they are sufficiently apart.

Fluctuation in the intensity of a downlink (satellite-to-ground) laser beam is examined by first measuring the acquisition time and pointing error of a satellite-borne laser communication system. The downlink intensity fluctuation is then simulated using a computer model that takes into account the effects of acquisition and tracking control, satellite attitude perturbation, and intensity fluctuation due to atmospheric turbulence. The simulation shows that when the tracking control-loop is closed, the major portion of the angular variation due to satellite attitude variation is corrected by the tracking control; scarcely any angular error in beam pointing remains.

The purpose of space surveillance is to classify and, if possible, assess the mission and performance capabilities of space objects. Historically, imaging techniques have obtained useful results. However, with the advances achieved in microtechnologies, small but highly functional satellites (largest dimension <1 m) are emerging that are hard to identify by imaging with large ground-based telescopes. The concept of using nonimaging measurements to obtain information is relatively new. In this paper, we present and discuss the performance of two techniques for classifying satellites based on spectral measurements. A distance-based classifier and a neural-net-based classifier are used to process both calibrated spectral data and features computed using these data. Neural networks are found to give better recognition results than the distance-based classifier, and once trained, this method is also faster. The average error rates for the distance-based method are greater than 30% when the inputs are the calibrated spectra, and 70% when using the central moments and the K-nearest-neighbors method. The best results are obtained for the neural network design, with the lowest class error rate at 0% for some satellites, the highest error rate at 30%, and an average error rate at 16%.

This paper presents a new blind-pixel correction and approximation algorithm based on moving-scene analysis for an infrared focal plane array. It is known that there exists strong correlation among neighboring images of an image sequence. By using moving-scene analysis, movement parameters can be estimated from an image sequence. Because the pixels in the previous frame will have strong correlation with the corresponding blind pixels in the current frame, the former can be used to correct or approximate the latter. Our proposed algorithm also can preserve the edges of target objects in the corrected image accurately. Theoretical analysis and experimental results have shown its effectiveness.

We present numerical calculations of the Fresnel diffraction for periodic structures in an optical system with two apertures. In such a system, measurements are affected by the relation between the spatial frequency of the sample and the geometrical parameters involved (i.e., aperture diameters, radiometer-sample distance, in-plane rotation, and translation of the sample). This numerical calculation of the Fresnel diffraction enables us to establish criteria to choose the right geometrical parameters of the system to ensure invariance of the measurements when the sample is rotated or shifted. We use the theory of partial coherence to calculate the Fresnel diffraction through two successive apertures. By using the point spread function of the system, as in the theory of partial coherence, we avoid complicated statistical processes that are commonly used in this theory. We show some numerical results that verify our proposal.

A dynamic model of IR missile engagements must combine the output of signature models into a scene of given resolution with a changing viewpoint and moving targets against some background. Some signature prediction models are stand-alone software packages, which currently cannot be dynamically interfaced to a running engagement model. They can be used to conveniently provide an image of an IR target at high resolution at a single viewpoint. An algorithm to solve the radiance map composition problem and combine target, sea, and sky signatures can be thought of intuitively as an idealized imaging radiometer model. We present the derivation of such a model from physical and signal processing considerations and its practical implementation; show that under reasonable assumptions, the radiance integral takes the form of a particularly simple computation; and demonstrate that the model behaves reasonably with translucent targets. The derived methodology provides very high radiometric accuracy with a rigorously controlled error and smooth integration of objects moving through the scene.

The standard antialiasing techniques within commercial graphics hardware are unsatisfactory for simulations involving targets at long ranges, e.g., that for imaging IR weapons. In this case, due to the presence of high-spatial-frequency components beyond the Nyquist frequency, the resulting scenes will contain aliasing and scintillation artifacts. Custom antialiasing techniques (that operate by supersampling) have been devised to deal with this; for example, zoom antialiasing and the corrected supersampling and scaling derivative. An alternative technique in which the target is prefiltered, shown to be equivalent to 3-D blurring of target objects at the vertex level, is described. An analysis of antialiasing performance is provided together with example imagery.

This paper presents an image restoration technique incorporating local statistical knowledge in the cost function. Instead of using a conventional grayscale-based error measurement such as the mean squared error, we compare local statistical information about regions in two images using a new error measure. Transient features such as edges and textures are more strongly emphasized than relatively homogeneous regions. With the addition of this local information, we attempt to provide a measure closer to human visual appraisal. We then extend the popular constrained squared-error cost function by incorporating this image error measure. Due to its nonlinear nature, conventional restoration algorithms cannot optimize this cost function efficiently. Therefore we seek an iterative approach. In particular, an extended neural network algorithm is proposed to perform the restoration. It is shown that this technique is efficient, effective, and robust. It compares favorably with other techniques when applied to both grayscale and color images. The results of a subjective survey comparing the proposed algorithm with a more conventional neural network algorithm are presented. The subjects tested in the survey overwhelmingly favored the results provided by the proposed method.

This paper addresses a novel robust watermarking method for digital images using local invariant features. Most previous watermarking algorithms are unable to resist geometric distortions that desynchronize the location where copyright information is inserted. We propose a watermarking method that is robust to geometric distortions. In order to resist geometric distortions, we use a local invariant feature of the image called the scale-invariant feature transform (SIFT), which is invariant to translation and scaling distortions. The watermark is inserted into the circular patches generated by the SIFT. Rotation invariance is achieved using the translation property of the polar-mapped circular patches. Our method belongs to the blind watermark, because we do not need the original image during detection. We have performed an intensive simulation to show the robustness of the proposed method. The simulation results support the contention that our method is robust against geometric distortion attacks as well as signal-processing attacks. We have compared our results with those of other methods, and our method outperforms them.

The limitations of electronics in conducting parallel arithmetic, algebraic, and logic processing are well known. Very high-speed (terahertz) performance cannot be expected in conventional electronic mechanisms. To achieve such performance we can introduce optics instead of electronics for information processing, computing, and data handling. Nonlinear optical material (NOM) is a successful candidate in this regard to play a major role in the domain of optically controlled switching systems. The character of some NOMs is such as to reflect the probe beam in the presence of two read beams (or pump beams) exciting the material from opposite directions, using the principle of four-wave mixing. In image processing, edge extraction from an image is an important and essential task. Several optical methods of digital image processing are used for properly evaluating the image edges. We propose here a new method of image edge detection, extraction, and enhancement by use of AND-based switching operations with NOM. In this process we have used the optically inverted image of a supplied image. This can be obtained by the EXOR switching operation of the NOM.

Efficient line-based very large scale integration architectures for the 2-D discrete wavelet transform (DWT) based on a lifting scheme, using the 9/7 wavelet filters adopted in the JPEG 2000 proposal, are proposed. The embedded decimation technique based on folding and time multiplexing was exploited to optimize the architecture, which reduces the size of buffer memory required and the amount of RAM access, and hence the occupied area and power consumption of the devices. Using this technique, a single-input, single-output architecture (SISOA) and a two-input, two-output architecture (TITOA) are proposed. The presented SISOA is designed to generate one output per clock cycle; the TITOA is designed to generate two outputs per clock cycle with the same memory requirement as that for SISOA, where the four subband coefficients of the transformed signal are available interleaved. Because only one line of data is required at a time, a single-port memory can be used. Performance analysis and comparison results demonstrate that the proposed method is economical of hardware cost and computation time. The advantages of the design also include short output latency, simple data flow, regularity, and scalability, as well as suitability for VLSI implementation.

A snake-based algorithm for segmenting an object from a pair of stereo images is presented. Unlike previously developed snake-based algorithms, this one performs well even when the objects in the picture are occluded and the background behind them is cluttered. Also, the algorithm is not sensitive to the placement of the initial snake points. The algorithm uses a new energy function defined over the disparity space between the pair of the stereo images to successfully locate the boundary of an object in a complex image. Experimental results have shown that the developed algorithm produces more accurate segmentation results than those of the well-known conventional snake algorithm reported by Kass et al.

Detection and outlining of boundaries of organs and tumors in computed tomography (CT) and magnetic resonance imaging (MRI) images are prerequisite in medical applications. A special design Hopfield neural network called the contextual Hopfield neural network (CHNN) is presented for finding the edges of CT and MRI images. Different from the conventional 2-D Hopfield neural networks, the CHNN maps the 2-D Hopfield network at the original image plane. With the direct mapping, the network is capable of incorporating pixels' contextual information into an edge-detecting procedure. As a result, the effect of tiny details and noise will be effectively removed by the CHNN. Furthermore, the problem of satisfying strong constraints can be alleviated and results in a fast converge. Our experimental results show that the CHNN can obtain more appropriate, more continued edge points than Laplacian-based, Marr-Hildreth's, Canny's, wavelet-based, and CHEFNN methods.

A one-dimensional approach to iris recognition is presented. It is translation-, rotation-, illumination-, and scale-invariant. Traditional iris recognition systems typically use a two-dimensional iris signature that requires circular rotation for pattern matching. The new approach uses the Du measure as a matching mechanism, and generates a set of the most probable matches (ranks) instead of only the best match. Since the method generates one-dimensional signatures that are rotation-invariant, the system could work with eyes that are tilted. Moreover, the system will work with less of the iris than commercial systems, and thus could enable partial-iris recognition. In addition, this system is more tolerant of noise. Finally, this method is simple to implement, and its computational complexity is relatively low.

We propose a video coding scheme to improve moving-target detection at very low bit rate, based on two key features: energy-based quantizer formation, and optimized interquantizer and intraquantizer prioritization. Rational Embedded Wavelet Video Coding (REVIC) is a fully implemented software video codec of low complexity and without motion compensated filtering to provide additional simplicity, adaptivity, and error resilience. It is shown to be quite effective in video coding of moving targets (e.g., military vehicles) at very low bit rates, while retaining the attributes of complete embeddedness for progressive transmission and scalability by fidelity and resolution. The proposed coding technique improves the explanatory power of decoded sequences (to achieve maximum target detection versus bit-rate performance) for a video compression system. The explanatory power of compressed sequences is important in surveillance applications, where trained video analysts may utilize decoded sequences to support decision processes in strategic, operational, and tactical tasks.

We propose a fast disparity estimation algorithm using background registration and object segmentation for stereo sequences from fixed cameras. Dense background disparity information is calculated in an initialization step, so that only disparities of moving object regions are updated in the main process. We propose a real-time segmentation technique using background subtraction and interframe differences, and a hierarchical disparity estimation using a region-dividing technique and shape-adaptive matching windows. Experimental results show that the proposed algorithm provides accurate disparity vector fields with an average processing speed of 15 frames/s for 320×240 stereo sequences on an ordinary PC.

A tapered spot-size converter with adjustable step length in two-dimensional photonic crystals has been proposed by removing several rows of dielectric rods from the photonic crystal structure. Its functionality and performance have been numerically investigated and simulated by the finite-difference time-domain method. The device has a conversion ratio of 10:1 and more than 80% power transmission efficiency for an optical communication wavelength of 1.55 µm. The converter has high compactness and potential applications as a connector in the optical signal connection between photonic-crystal-based waveguide devices and conventional micro-optical waveguide devices or single-mode optical fibers.

Propagation of higher order annular Gaussian (HOAG) laser beams in free space is examined. HOAG beams are defined as the difference of two Hermite-Gaussian (HG) beams; thus, they can be produced by subtracting a smaller beam from a larger beam, that are cocentered and both possess HG mode field distributions. Such beams can be considered as a generalization of the well-known annular Gaussian beams. We formulate the source and receiver plane characteristics and kurtosis parameter of HOAG beams propagating in free space and evaluate them numerically. In comparison to HG beams, HOAG beams have a broader beam size with outer lobes of kidney shape. The amount of received power within the same receiver aperture size, that is, power in bucket, is generally lower for higher order beams. The convergence of the kurtosis parameter to an asymptotic value for higher order beams takes much longer propagation distances compared to zero-order beams.