This PDF file contains the editorial “The Tower of Google I” for OE Vol. 45 Issue 01

A master-oscillator fiber power amplifier (MOPA) system with a 4-m-long Yb³⁺-doped homemade large mode area (LMA) double-clad fiber is reported. The system emits up to 133.8 W of amplified radiation at a wavelength of 1064 nm and a repetition rate of 100 kHz, limited only by the available pump power. Peak power of 300 kW at 20 kHz with a pulse duration of 15 ns is obtained.

MONS (Measuring Oscillations in Nearby Stars) is a 320-mm-aperture telescope designed as the primary payload of the proposed Danish Small Satellite Program's Rømer mission. If launched, MONS would observe the oscillations in about 20 stars by means of highly sensitive photometric measurements of the relative blue and red signal levels, enabling new insights into the structure and evolution of solar-type stars. We describe the key aspects of the telescope design and analysis of the optical design.

The acquisition of the colorimetric information about an object using a color image acquisition device is important at an early stage in a color management system. The accuracy of the colorimetric values estimated by the device responses depends not only on the spectral sensitivities of a set of sensors but also on the noise present in the devices. We address the optimization of a set of spectral sensitivities with Gaussian distribution functions based on a colorimetric evaluation model. It is demonstrated that the design of optimal sensors is contingent on finding the right balance between the human visual subspace and the subspace that maximizes the singular values of a matrix SLVΛ^1/2 to increase the robustness to noise, where S, L, V, and Λ represent a sensor matrix, a diagonal matrix for an illuminant, a basis matrix, and a diagonal matrix with eigenvalues of an autocorrelation matrix of reflectance spectra, respectively.

An inspection system for the surface-profile properties of an organic photoconductor (OPC) drum substrate is studied. Defective substrates have approximately 1-mm-period waves in their surface profiles along the axial direction. The slope distribution of the surface profile is measured with an optical inspection system, which detects the angular deflection of a laser beam scanned over the surface at a high speed of 15 mm/ms. To discriminate between good and defective substrates, a threshold decision is made on components of the experimentally measured power spectrum of the slope distribution around 1-mm spatial period. The inspection system provides the same results as visual inspection with an accuracy better than 6σ. The setup dose not require any vibration isolators, because of its short inspection time of 2 ms.

A color phase-shifting technique has been recently developed for high-speed 3-D shape measurement. In this technique, three sinusoidal phase-shifted images used for a measurement cycle in a traditional grayscale phase-shifting technique are encoded into one color image. Therefore, only a single color image is needed for reconstructing the 3-D surface shape of an object. The measurement speed can then be increased up to the frame rate of the camera. However, previous experimental results showed that the measurement accuracy of this technique was initially low, due largely to the coupling and imbalance of color channels. In this paper, two solutions, one software-based and one hardware-based, are proposed to compensate for these errors. Experimental results show that the second solution—modification of the camera together with an imbalance compensation algorithm—would effectively reduce the errors and produce better measurement results than the software-based compensation method. This technique has many potential applications in high-speed measurement, such as highway inspection and dynamic measurement of human body.

A new pixel structure with twice-bent beams and eaves structure, suitable for high-resolution uncooled infrared (IR) focal plane arrays (FPAs), is proposed. In comparison with previous results (FPA of 37-µm pixel pitch), the thermal conductance of the test device with the proposed pixel structure of 23.5-µm pitch is reduced about 2.5 times. The eaves structure, which is adopted to increase the fill factor of pixels, improves the responsivity by a factor of 1.3. A 640×480 bolometer-type uncooled IRFPA is demonstrated by utilizing the new pixel structure, with supplementary modification to improve thermal conductance and thermal time constant. It shows a noise equivalent temperature difference (NETD) of 50 mK for F/1.0 optics at 30 frames/sec, a thermal conductance of 0.03 µW/K, and a thermal time constant of 16 msec.

Oxide vertical-cavity surface-emitting lasers (VCSELs) are widely used in high-speed fiber optical communications. Burn-in tests are run on oxide VCSELs to weed out defective devices, thereby fulfilling the related reliability requirements. Nevertheless, some oxide VCSELs fail in the field. Oxide VCSELs that failed in the field were examined with transmission electron microscopy (TEM). Besides cross-sectional views, plan views of the entire active region were observed. Development of a dislocation network in the active region was found in all samples, and that is the cause of the failures. The Burgers vectors of these dislocations are parallel to [101]. Although the trigger of the dislocations has not been identified, electrostatic discharge (ESD) does not cause the dislocation in our investigated devices, since no ESD-damaged region was found. Strain and/or crystal defects introduced during VCSEL fabrication are considered to be the cause.

Coherent laser light was tuned by cumulative refraction of light waves in a semiconductor laser cavity, while the wavelength of electroluminescence was not tuned by any noticeable amount. This effect is explained by the fact that the laser light has perfect spatial coherence, while the electroluminescence has a spatial coherence of ≈ 0.7 µm, smaller than the width of the wavefront. The refraction angle of 2.4 deg means that more than 16 triangular electrodes are involved in the refraction. This observation of coherent tuning may lead to a new optoelectronic device, a coherent filter, which filters out only coherent light from a coherent-incoherent mixture.

Laser drilling is very important in many industries such as automotive, aerospace, electronics, and materials processing. It can be used to produce critical components with novel hole geometry for advanced systems. Percussion drilling and trepanning are two laser drilling methods. In the conventional trepanning method, a laser beam is scanned along a circular or spiral orbit to remove material to achieve a desired hole shape. These orbits generally trace a circular path at the inner wall of the holes. This suggests that an annular beam can be used to accomplish trepanning, a technique we refer to as optical trepanning. The ray-tracing technique of geometrical optics is employed in this paper to design the necessary optics to transform a Gaussian laser beam into an annular beam of different intensity profiles. Such profiles include uniform intensity within the annulus, full Gaussian with maximum intensity, and half Gaussian with maximum intensities at the inner and outer surfaces of the annulus.

A bending-sensitive fiber (BSF) is fabricated and analyzed to create a microbend fiber sensor (MFS) with a simple structure. The BSF exhibits a gradual rather than sudden change in the microbend loss, according to a variation in the microbend applied to the BSF. According to the measured refractive index profile of the fabricated BSF, the proposed BSF consists of three different regions: the first core equal to the core of a SMF; the second core that is located around the first core, and whose refractive index is lower than the first core and higher than the cladding; and the cladding. The 3-D finite difference beam propagation method (3-D FD-BPM) is utilized to analyze the characteristics of the BSF. Based on the numerical results using 3-D FD-BPM and the fabricated BSF, a simple MFS with a BSF is created and compared to the MFS with a SMF. In particular, the MFS with the BSF showed a microbend loss from –1 to –20 dB at 1550 nm when the pressure given to the optical fiber is varied from 0 to 0.05 MPa; meanwhile, the MFS with the SMF showed no optical power attenuation.

The fabrication and testing results of a 65-pin ceramic packaged 4×4 arrayed position sensitive detector are presented. The detector, consisting of 16 tetra-lateral sensitive areas, is a p-n-n+ configuration made on 3-in. <111> n-type high resistance crystal silicon substrates. A 100-nm antireflection SiO₂ thin film is formed on the surface, along with a multilayer cover glass with transmissivity >98% from 400 to 950 nm. Primary tests of the device show that it has a low dark current, high spectral sensitivity, very fast response speed, and very good linearity. The dark current of an element unit is less than 20 nA, which is the allowable maximum dark current. The peak spectral sensitivity of the sensor is over 505 mA/W at 800-nm wavelength. Its response time is 8 ns at 45-V reverse bias and the nonlinearity of the total sensitive area is less than 1%.

We propose a new and simple method to reconstruct the refractive-index modulation of a symmetric fiber Bragg grating (FBG) from its magnitude reflection spectrum (i.e., phase information is not required). The reconstruction method uses an FBG model based on some known parameters of the fabricated FBG (grating length, number of subgratings, and grating period) and an optimization technique, namely, the Quasi-Newton method, to find the local parameter values of the FBG. To our knowledge, this is the first time that the reconstruction of FBG is experimentally realized by an optimization technique. Compared with other reconstruction methods, the proposed method is so simple that it requires only the reflection spectrum of the fabricated FBG.

The performance of InGaN LEDs in terms of photon extraction efficiency is analyzed by the Monte Carlo photon simulation method. Simulation results show that the sidewall slanting scheme, which works well for the AlInGaP or InGaN/SiC system, plays a very minimal role in InGaN/sapphire systems. In contrast to InGaN/SiC systems, a lower refractive index sapphire substrate restricts the generated photons to enter the substrate, minimizing the chances for the photons to be deflected by the slanted sidewalls of the epitaxial semiconductor layers that are usually very thin. The limited photon transmission to the sapphire substrate also degrades the photon extraction efficiency, especially in the epitaxial side down mount. One approach to exploit the photon extraction potential of the epitaxial side down mount may be to texture the substrate-epitaxy interface, by possibly growing the epitaxial layers on a sapphire substrate that is either appropriately surface textured or patterned and etched. In this case, randomized photon deflection off the textured interface directly increases the number of photons entering the sapphire substrate, from which they easily couple out of the chip, thereby improving the photon extraction efficiency drastically.

A large-scale optical matrix switch is generally composed of a permutation of switching units connected by a network of optical signal paths. In this work, a configuration of optical matrix switches with flexible switching units and Banyan networks is proposed. With Banyan networks, silica-based waveguides, and two types of switching units, the insertion loss of the optical matrix switch configuration is extensively studied for serial scales from 2×2 to 64×64. Typically for an 8×8 optical matrix switch, the insertion losses are achieved 3.0 dB for a rearrangeably nonblocking configuration with two-state 2×2 switching units, and 4.7 dB for a strictly nonblocking configuration with three-state 2×2 switching units when the intersection angle is designed as 30 deg. Even at the released intersection angle of 20 deg, the insertion losses of 4.0 and 5.8 dB can be achieved for the rearrangeably nonblocking and strictly nonblocking configurations, respectively. In addition, the blocking property and the scalability of the large-scale matrix optical switches with this regime are analyzed.

We propose a novel Z-cut lithium niobate electro-optic modulator based on a Si₂N₂O substrate. The structure is analyzed using the finite element method. The novel modulator with 3-dB optical bandwidth, half-wave voltage, and characteristic impedance at 1.5-µm wavelength are 120 GHz, 3.5 V and 50.2Ω, respectively. The simulation results show that the modulator can achieve wide bandwidth, low half-wave voltage, and a good impedance match. The modulator has good potential for high-speed optical transmission systems.

A new method is presented to perform bit synchronization and detection of binary nonreturn-to-zero (NRZ) data from a free-space optical (FSO) communication link. Based on the wavelet transformation, a new bandpass filter is developed and implemented. It is shown that the Haar wavelet is an excellent choice for this purpose. The center frequency of this filter is a function of the scale and could be adjusted to adapt to the variation of the channel. The output of the filter is zero mean and is closely related to the derivative of the binary data. The filter has a linear phase; therefore, its output is used for synchronization and detection of the data. Analysis of the method is presented using Fourier transformation. In addition, adaptive Wiener filtering is utilized to reduce the effect of the additive white Gaussian noise in the data. Simulation experiments are performed and presented using real and synthetic data. The results of the experiments indicate that the Haar wavelet transform and adaptive Wiener filtering are robust and effective tools in dealing with FSO data.

We investigate the performance of wavelength shift keying (WSK) technique and repeated unequal channel spacing (RUS) scheme and compare them considering a repeaterless optical wavelength-division-multiplexed (WDM) transmission system in the context of four-wave mixing (FWM) effect. The WSK-WDM system outperforms the RUS-WDM system within particular signal levels and enables higher allowable input power, lower power penalty, and higher transmission distance for a given bit error rate at the expense of increased bandwidth. The comparative performance provides important design considerations for a long-haul optical WDM system.

We propose and evaluate two new schemes for providing differentiated services in optical burst switched (OBS) networks. The two new schemes are suitable for implementation in OBS networks using just-in-time (JIT) or just-enough-time (JET) scheduling protocols. The first scheme adjusts the size of the search space for a free wavelength based on the priority level of the burst. A simple equation is used to divide the search spectrum into two parts: a base part and an adjustable part. The size of the adjustable part increases as the priority of the burst becomes higher. The scheme is very easy to implement and does not demand any major software or hardware resources in optical cross-connects. The second scheme reduces the dropping probability of bursts with higher priorities through the use of different proactive discarding rates in the network access station (NAS) of the source node. Our extensive simulation tests using JIT show that both schemes are capable of providing tangible quality of service (QoS) differentiation without negatively impacting the throughput of OBS networks.

An all-fiber interferometric experimental technique to determine the mode effective index n_eff of a single-mode fiber is proposed. The method involves setting up an unbalanced Mach-Zehnder interferometer (MZI) with the test fiber. By inducing a small change in the differential path length between the two arms of the MZI, a proportional change in the periodicity of its transmission spectrum could be introduced. A precise measurement of this change in periodicity has been used to accurately determine the value of n_eff of the test fiber.

We propose a novel, continuously tunable, photonic microwave notch filter configuration that uses, for the first time, single-sideband (SSB) modulated optical signals together with a linearly chirped fiber grating in a Sagnac loop. Measured results are presented to show a large notch rejection and an easily tunable free spectral range (FSR) by tuning the chirped fiber Bragg grating (FBG). The configuration enables doubling of time delay in the Sagnac loop. Use of SSB offers the potential to overcome chromatic dispersion.

A full-duplex radio-on-fiber (ROF) transport system based on wavelength division multiplexing (WDM) and optical add-drop multiplexing techniques, as well as employing a simple fiber Bragg grating (FBG) filter at the receiving site, is proposed and demonstrated. Over a long-haul fiber link, good transmission performance is obtained in our proposed systems. Since our proposed system does not use a sophisticated optical single sideband (SSB) transmitter, it reveals a prominent alternative with advantages in simplicity and cost. Such a proposed full-duplex ROF transport system is suitable for the long-haul microwave optical link.

The coupling characteristics of newly proposed hollow optical fiber (HOF) couplers are rigorously analyzed using the compact 2-D finite-difference time-domain (FDTD) method. The FDTD-simulated coupling efficiency as a function of the coupling length, and the transverse electric field distributions of the higher-order mode converted by the HOF couplers, are investigated. We propose that the designed HOF couplers can be used for various optical devices such as side-pumping double-clad fibers, mode splitters, and power monitors.

In high-speed computation, optics has established itself as a successful candidate. Many proposals have already been adopted for performing arithmetic, algebraic, and logic operations using light as the information carrier. We propose a new method of optical data comparison with the proper use of nonlinear material applying the 1's compliment method. The binary optical data are manipulated here properly to achieve the required result.

Full-field optical coherence tomography (OCT) using a complementary metal-oxide semiconductor (CMOS) camera with an integrated a digital signal processor (DSP) is demonstrated. The CMOS-DSP camera employed is typically used in machine vision applications and is based on an array of 1024×1024 direct readout pixels that are randomly addressable in space and time. These characteristics enable the camera to be used as a fast full-field detector in carrier-based optical metrology systems. The integrated DSP facilitates basic signal processing including real-time filtering and undersampling. The optical setup used to implement this OCT method is composed of a free-space Michelson interferometer and a superluminescent diode (SLD) light source, with an electromechanical shaker for depth scanning. Unlike classical OCT approaches, however, the setup does not require any electromechanical device for lateral scanning. A 64×30 pixel region of interest was imaged at 235 frames/s and sampled in depth, corresponding to a volumetric measurement of 875×410×150 µm. Measurements carried out on a simple calibration specimen indicated lateral and axial resolutions of 14 and 22 µm, respectively. The presented approach offers an inexpensive and versatile alternative to traditional OCT systems and provides the basis for a functional machine vision system suitable for industrial applications.

Multifrequency digital holographic reconstructions of primary sound sources embedded in scattering media are demonstrated. The sound field is measured with a scanning laser Doppler vibrometer (LDV) and broadening of the spectral content of the sound source is achieved by tuning the primary ultrasound (US) transducer around its resonance frequency. The results show that combining the complex amplitudes from the different frequency reconstructions results in a reduced susceptibility to multiple scattered sound and makes possible a quite thorough localization of the primary sound source. The depth resolution obtained is 11 US frequencies. This depth sensitivity is improved even further to only 2.8 wavelengths by applying a filter determined from the standard deviation over the phases.

A detailed model of the diffraction of plane and Gaussian beams on plane uniform phase Bragg gratings based on Kogelnik's theory of coupled waves is presented. The model describes transmitting gratings with arbitrary orientation in a plane-parallel plate taking into account spectral width and angular divergence of laser beams along with material dispersion of a photosensitive medium. The model results are compared with experimental data for high-efficiency Bragg gratings in a photothermorefractive (PTR) glass.

Atmospheric scintillation effects on point targets and small targets are studied in the middle-wave infrared (MWIR) band over a 2.5-km horizontal path. Expressions are presented for both exact numerical calculations and for a Gaussian-aperture approximation of scintillation that has an accurate analytic solution. Experimental results show that measurements of C_n² using a commercial NIR scintillometer allow the prediction of average scintillation levels in the MWIR when corrections are made for wavelength and dimensions of the source and receiver. The measured log-irradiance values are approximately lognormal, and are about a factor of 2 higher than would have been expected from theory. Our results are relevant for analysis of detection and false-alarm probabilities over long horizontal paths in the MWIR.

Band division is an important basis in radiative calculations, and the configuration of band divisions for various research purposes directly influences the accuracy and speed of radiative transfer computations. We explore four band-division schemes and their impacts on computed radiative fluxes and cooling rates. We explain that discrepancies in solar radiation at the surface that exist between radiation models and observations under clear-sky conditions arise partly from ignoring minor gases and weak absorption bands for major gases.

Spectroscopic measurements of infrared molecular transitions in gas plumes are evaluated for their potential to yield a reliable remote sensing technique for determination of plume temperature. Under optically thin conditions, the ratio of intensities from two different CO₂ transitions has no dependence on molecular concentration, but has sufficient thermal sensitivity to be used as a temperature diagnostic. Fine spectral resolution is not required. Experimental results involving simultaneous thermocouple and spectroscopic measurements support the use of spectroscopic intensity ratios to determine stack plume temperatures. Measurements involve a plume from a vehicle exhaust and a stack plume from a propane-burning portable plume generator. Observed CO₂ emission near 4.26 µm is affected by optical thickness and by self-absorption of emitted radiation by the cool outer regions of the plume. Measurements on the wings of molecular bands mitigate these absorption effects. Relationships between absorptive and emissive spectral intensities are developed that allow calibration of the temperature diagnostic using readily available absorption spectra. Based on data spanning a range of plume temperatures from 310 to 606 K, the root-mean-square difference between the spectroscopically derived temperatures and thermocouple data is 22 K.

Methane seepage is indicative of petroleum or natural gas reserves. Techniques aimed at detecting methane seepage with surface-based instrumentation have progressed significantly in recent years. These techniques rely on measurement of light attenuation due to methane absorption of short wave infrared (SWIR) radiation. Detection of methane seepage over water bodies with electro-optical remote sensing has been limited by the low surface reflectance of water. Also, due to sensor saturation, imagery over sunglint is commonly discarded in satellite remote sensing, because the glint conditions produce high surface reflectance. However, recent measurements in the SWIR of sunglint regions have revealed that the surface reflectance is spectrally flat and enhanced without causing saturation. This higher surface reflectance in sunglint regions allow for retrieval of the total column methane amount using ratios of measured radiances at wavelengths inside and outside the methane SWIR-absorbing channels. The methane retrieval method presented here, based on shortwave infrared band ratios in sunglint regions, allows for detection of methane seepage over the Earth's oceans and lakes, and the detection of possible petroleum or natural gas reserves. Radiative transfer simulations are used to demonstrate the capabilities offered by this technique.

Recent advances in uncooled infrared technology have resulted in thermal imagers with resolution approaching that of cooled counterparts at a significantly lower cost. We investigate the application of linear classification schemes to a database consisting of 420 images collected from 14 adult subjects using an uncooled infrared camera under indoor controlled conditions. Results show that the linear discriminant approach (LDA) leads to the best classification performances (99.3%), while the best principal component analysis (PCA)-based scheme leads to an accuracy of 91.33%. Results also show that PCA-based classification scheme performance improves by removing the top three eigenvectors, associated with the three largest eigenvalues, from consideration in the generation of the PCA projection matrix for the small database considered in this study, as was noted in visible imaging face recognition studies.

An in-depth analysis is made in the case of object degeneration and ghosting artifacts in a neural-network-based nonuniformity correction algorithm (NN-NUC) for infrared focal plane arrays (IRFPAs). It is found that updating the correction coefficients blindly in the NN-NUC scheme without taking the object edge into account is the root of the problem. Based on this conclusion, an edge-directed NN-NUC scheme (ED-NN-NUC) is proposed to eliminate ghosting artifacts and object degeneration. Comparison experiments with simulated data and real IRFPA infrared data show that the root of the problem pointed out is correct and the proposed scheme is rational and effective.

A multimedia database system should deal efficiently with both image compression and retrieval functions. It is critical to develop image indexing techniques that search databases based on their content in a compressed domain. We propose a new scheme, query by index image, based on vector quantization, to facilitate image retrieval in a compressed domain. The proposed algorithm exploits different index images obtained by sorting codevectors to capture various kinds of image feature. Hence, intrablock correlation and interblock correlation in an image can be efficiently represented. Our proposed algorithm not only can extract features from the pixel domain but also from a transform domain, such as that of wavelet coefficients. Experimental results demonstrate that the retrieval performance of the proposed scheme is more accurate than that of other similar methods.

Compared to conventional video standards, the main features of H.264 standard are its high coding efficiency and its network friendliness. In spite of these outstanding merits, it is not easy to implement H.264 codec as a real-time system, due to its requirements of large memory bandwidth and intensive computation. Although the variable-block-size motion compensation using multiple reference frames is one of the key coding tools to bring about its main performance gain, its optimal use demands substantial computation for the rate-distortion calculation of all possible combinations of coding modes and estimation of the best motion vector. Many existing fast motion estimation algorithms are not suitable for H.264, which employs variable motion block sizes. We propose an adaptive motion search scheme utilizing the hierarchical block structure based on the deviation of subblock motion vectors. The proposed fast scheme adjusts the search center and search pattern according to the subblock motion-vector distribution.

We propose a method for feature-preserving regularized reconstruction in coherent imaging systems. In our framework, image formation from measured data is achieved through the minimization of a cost functional, designed to suppress noise artifacts while preserving features such as object boundaries in the reconstruction. The cost functional includes nonquadratic regularizing constraints. Our formulation effectively deals with the complex-valued and potentially random-phase nature of the scattered field, which is inherent in many coherent systems. We solve the challenging optimization problems posed in our framework by developing and using an extension of half-quadratic regularization methods. We present experimental results from three coherent imaging applications: digital holography, synthetic aperture radar, and ultrasound imaging. The proposed technique produces images where coherent speckle artifacts are effectively suppressed, and important features of the underlying scenes are preserved.

We propose a new method for an objective measurement of video quality. By analyzing subjective scores of various video sequences, we find that the human visual system is particularly sensitive to degradation around edges. In other words, when edge areas of a video sequence are degraded, evaluators tend to give low quality scores to the video, even though the overall mean squared error is not large. Based on this observation, we propose an objective video quality measurement method based on degradation around edges. In the proposed method, we first apply an edge detection algorithm to videos and locate edge areas. Then, we measure degradation of those edge areas by computing mean squared errors and use it as a video quality metric after some postprocessing. Experiments show that the proposed method significantly outperforms the conventional peak signal-to-noise ratio (PSNR). This method was also independently evaluated by independent laboratory groups in the Video Quality Experts Group (VQEG) Phase 2 test. The method consistently provided good performances. As a result, the method was included in international recommendations for objective video quality measurement.

The problems with binary watermarking schemes are that they have only a small amount of embeddable space and are not robust enough. We develop a slice-based large-cluster algorithm (SBLCA) to construct a robust watermarking scheme for binary images. In SBLCA, a small-amount cluster selection (SACS) strategy is used to search for a feasible slice in a large-cluster flappable-pixel decision (LCFPD) method, which is used to search for the best location for concealing a secret bit from a selected slice. This method has four major advantages over the others: (a) SBLCA has a simple and effective decision function to select appropriate concealment locations, (b) SBLCA utilizes a blind watermarking scheme without the original image in the watermark extracting process, (c) SBLCA uses slice-based shuffling capability to transfer the regular image into a hash state without remembering the state before shuffling, and finally, (d) SBLCA has enough embeddable space that every 64 pixels could accommodate a secret bit of the binary image. Furthermore, empirical results on test images reveal that our approach is a robust watermarking scheme for binary images.

A method to design an input phase mask for double-random phase-encoding optical encryption is proposed. The input phase mask is iteratively designed so that the extent of the Fourier spectrum of the product of an input image and the input phase mask correspond to the space bandwidth of the optical system. By using the designed input phase mask, we show that the bit-error rate is improved. We also discuss the number of phase levels of the designed input phase mask.

In manufacturing a lenticular display system, precise alignment of the lenticular sheet on the LCD panel may not be practically achievable. Hence, observed view images inevitably produce unwanted distortion. We propose a novel method to alleviate the display distortion of each observed view image in a lenticular 3-D display. We first derive the relationship between subpixel values on the LCD pixel array and the image to be observed at each viewing zone in terms of system design parameters and the viewer's eye position. Based on this relationship, we analyze the distortion between the observed and original view images. We then derive a compensation algorithm to minimize the distortion and generate high-quality 3-D images. To verify the proposed scheme, we examine displayed results from several 3-D images of synthetic and real scenes. The results demonstrate that the proposed scheme significantly reduces distortions and improves the image quality in the lenticular display system.

We investigate the application of independent-component analysis (ICA) to remotely sensed hyperspectral image classification. We focus on the performance of two well-known and frequently used ICA algorithms: joint approximate diagonalization of eigenmatrices (JADE) and FastICA; but the proposed method is applicable to other ICA algorithms. The major advantage of using ICA is its ability to classify objects with unknown spectral signatures in an unknown image scene, i.e., unsupervised classification. However, ICA suffers from computational expensiveness, which limits its application to high-dimensional data analysis. In order to make it applicable or reduce the computation time in hyperspectral image classification, a data-preprocessing procedure is employed to reduce the data dimensionality. Instead of using principal-component analysis (PCA), a noise-adjusted principal-components (NAPC) transform is employed for this purpose, which can reorganize the original data with respect to the signal-to-noise ratio, a more appropriate image-ranking criterion than variance in PCA. The experimental results demonstrate that the major principal components from the NAPC transform can better maintain the object information in the original data than those from PCA. As a result, an ICA algorithm can provide better object classification.

Extracting a smooth curve from unordered data has many applications to image analysis. However, many reported methods assume either that the shape of the input data is known a priori or that the boundary of the data is clearly defined. We present a method that can handle several types of data sets. The main idea of the method is to extract a generalized curve, which passes through the data set. The proposed method is able to extract a smooth curve from complicated unordered pattern data and without any prior knowledge of the shape of the input data. Experimental results show that our method can produce good results for many data sets including handwritten Chinese characters.

Novel view synthesis (NVS) is an important problem in image rendering. It involves synthesizing an image of a scene at any specified (novel) viewpoint, given some images of the scene at a few sample viewpoints. The general understanding is that the solution should bypass explicit 3-D reconstruction of the scene. As it is, the problem has a natural tie to interpolation, despite that mainstream efforts on the problem have been adopting formulations otherwise. Interpolation is about finding the output of a function f(x) for any specified input x, given a few input-output pairs {(x_i,f_i):i=1,2,3,...,n} of the function. If the input x is the viewpoint, and f(x) is the image, the interpolation problem becomes exactly NVS. We treat the NVS problem using the interpolation formulation. In particular, we adopt the example-based everything or interpolation (EBI) mechanism—an established mechanism for interpolating or learning functions from examples. EBI has all the desirable properties of a good interpolation: all given input-output examples are satisfied exactly, and the interpolation is smooth with minimum oscillations between the examples. We point out that EBI, however, has difficulty in interpolating certain classes of functions, including the image function in the NVS problem. We propose an extension of the mechanism for overcoming the limitation. We also present how the extended interpolation mechanism could be used to synthesize images at novel viewpoints. Real image results show that the mechanism has promising performance, even with very few example images.