Search

A new method for real-time edge enhancement and image equalization using photochromic filters is presented. The reversible self-adaptive capacity of photochromic materials is used for creating an unsharp mask of the original image. This unsharp mask produces a kind of self filtering of the original image. Unlike the usual Fourier (coherent) image processing, the technique we propose can also be used with incoherent illumination. Validation experiments with Bacteriorhodopsin and photochromic glass are presented. OCIS codes: 100.1160, 100.2980.

A temporal filtering technique, complementary to spectral filtering, has been developed for laser-induced fluorescence measurements. The filter is applicable in cases where the laser-induced interfering signals and the signal of interest have different temporal characteristics. For the interfering-signal discrimination a picosecond laser system along with a fast time-gated intensified CCD camera were used. In order to demonstrate and evaluate the temporal filtering concept two measurement situations were investigated; one where toluene fluorescence was discriminated from interfering luminescence of an aluminum surface, and in the other one Mie scattering signals from a water aerosol were filtered out from acetone fluorescence images. A mathematical model was developed to simulate and evaluate the temporal filter for a general measurement situation based on pulsed-laser excitation together with time-gated detection. Using system parameters measured with a streak camera, the model was validated for LIF imaging of acetone vapor inside a water aerosol. The results show that the temporal filter is capable of efficient suppression of interfering signal contributions. The photophysical properties of several species commonly studied by LIF in combustion research have been listed and discussed to provide guidelines for optimum use of the technique. OCIS codes: 120.1740, 120.2440, 280.1740, 280.2490, 290.2648, 300.2530.

In this paper, a design method for a multiwavelength photon sieve is described based on a random-area-divided approach, where the whole aperture of a multiwavelength imaging photon sieve is divided into multiple discrete spaces corresponding to the number of the selected working wavelengths. The micropinhole distribution in each discrete space can be calculated for the defined wavelength with one fixed focal length in terms of the normal design for photon sieve. A three-wavelength photon sieve was designed and fabricated in the lab, and its imaging properties are analyzed in the experimental optical system with satisfactory results. OCIS codes: 050.1970, 220.4000, 350.3950, 110.4234.

Laser guide star wavefront sensing comes with several limitations. When imaged with a Shack-Hartmann wavefront sensor, the laser guide star is seen as extended sources elongated in the directions given by the lenslet locations and the laser axis. A test bed has been built in the Adaptive Optics Laboratory of the University of Victoria that reproduces this effect as seen on extremely large telescopes. A new wavefront sensing algorithm, the matched filter, has been implemented and its performance assessed with the test bed. Its ability to mitigate laser guide star aberrations by tracking the sodium layer fluctuations in a closed loop adaptive optics system is shown. OCIS codes: 000.3110, 010.0280, 010.1080, 010.3310, 010.7350.

A new approach to a polarization-independent twisted liquid-crystal (LC) structure, where phase difference between orthogonal eigenmodes is tuned to be an integer multiple of 2[pi], is demonstrated with a numerical model. For select wavelengths, polarization-independent operation can be achieved by tuning the twist rate and thickness of the LC cavity. Applications can be found in polarization- independent switches and field sequential wavelength selection devices. OCIS codes: 230.3720, 050.2230.

We extend the approach recently applied for achieving spatial and temporal superresolution to spectral measurements with superresolution. The light passes previously through a filter with a periodic spectral response. Owing to the spectral modulation, the parts of the spectrum Fourier transform are shifted and transmitted through the passband of the Fourier transform of the spectrometer instrumental function. Thus, all the parts of the spectrum Fourier transform are recorded by the spectrometer and then restored by a special procedure. An inverse Fourier transform gives the spectrum restored with superresolution. We numerically demonstrate more than tenfold enhancement of the resolution by using a sampled fiber Bragg grating for spectral modulation. OCIS codes: 300.6320, 060.3735, 070.2615, 070.4790.

In this paper, a new framework to filter speckle noise in polarimetric synthetic aperture radar (PolSAR) data is presented. The proposed filter, named the model-based PolSAR (MBPolSAR) filter, is based on exploiting the multiplicative--additive speckle noise model for multidimensional SAR data. The entries of the sample covariance matrix are processed according to this multiplicative--additive speckle noise model as a function of the complex correlation coefficient. Hence, the covariance matrix elements are processed differently. This filtering scheme improves the reduction of speckle noise and ameliorates the estimation of the polarimetric information. The filtering performances of the MBPolSAR approach are first tested quantitatively by means of simulated PolSAR data. In a second stage, they are evaluated by means of experimental SAR data acquired by the Deutsches Zentrum fur Luft-und Raumfahrt. Index Terms--Filtering, polarimetric radar, polarimetry, speckle, synthetic aperture radar (SAR).

Preece and Claridge [IEEE Trans. Pattern Anal. Mach. Intell. 26, 913 (2004)] have proposed a technique for selecting filters for the maximally accurate recovery of object parameters such as chromophore concentrations from a multispectral image of an object. Their selection criteria are derived from an analysis of a model of light propagation in the object and take into account both errors in the modeling process and errors in the image acquisition process, as well as the inherent behavior and structure of the model. We investigate their method on simulated image data and show that filters selected according to their criteria are demonstrably superior to other choices. OCIS codes: 110.4234, 100.3190, 170.3010, 170.4580, 170.6935.

Using metallic film perforated with a subwavelength periodic structure, a novel concept of a color filter for multicolor organic light-emitting diode (OLED) display is proposed. Based on the phase-matching condition for extraordinary optical transmission, three primary color emissions can be obtained by optimizing the structure's periodicity. Two periodic structures, an array of one-dimensional periodic slits and a two-dimensional periodic hole array, are studied using coupled mode theory. Also, the feasibility of applying these structures as color filters is analyzed. The relative intensity at the unwanted wavelength, which is generated by higher resonant transmission, had been calculated to eliminate its effect on the purity of these filters. It is important that this type of color filter simultaneously solves the low emission efficiency problem for OLEDs with the aid of enhanced transmission of metal film. OCIS codes: 230.7408, 310.6628, 310.6845.

We discuss and implement a log-polar transform-based distortion-invariant filter for automatic target recognition applications. The log-polar transform is a known space-invariant image representation used in several image vision systems to eliminate the effects of scale and rotation in an image. For in-plane rotation invariance and scale invariance, a log-polar transform-based filter was synthesized. In cases of in-plane rotation invariance, peaks shift horizontally, and in cases of scale invariance, peaks shift vertically. To achieve out-of-plane rotation invariance, log-polar images were used to train the wavelet-modified maximum average correlation height (WaveMACH) filter. The designed filters were implemented in the hybrid digital-optical correlation scheme. It was observed that, for a certain range of rotation and scale differences, the correlation signals merge with the strong dc. To solve this problem a shift was introduced in the log-polar image of the target. The use of a chirp function for dc removal has also been discussed. Correlation peak height and peak-to-sidelobe ratio have been calculated as metrics of goodness of the log-polar transform-based WaveMACH filter. Experimental results are presented. OCIS codes: 070.0070, 070.4550, 070.5010, 100.3008.

A novel multiline filter using a two-dimensional guided-mode resonant (GMR) filter is proposed. The filter concept utilizes the multiple planes of diffraction produced by the two-dimensional grating. Multiple resonances are obtained by matching the guided modes in the different planes of diffraction to different wavelengths. It is shown that the location and the separation between resonances can be specifically controlled by modifying the periodicity of the grating and the other physical dimensions of the structure. This is in contrast to the one-dimensional GMR filters where the location of the resonances is material dependent. Two-line reflection filter designs with spectral linewidths less than 1 nm and a controllable spectral separation of up to 23% of the short resonance wavelength are presented using rectangular-grid grating GMR structures. Three-line filters are designed in hexagonal-grid grating GMR structures with two independently controllable resonance locations. OCIS codes: 050.1950, 130.3120, 230.7370.

An empirical procedure based on optical-density-bandwidth products was recently proposed for thickness estimation of dielectric thin film reflectors. A parallel is established with new results derived from the Fourier transform thin film synthesis technique. Two Fourier-transform approaches are proposed and justified by numerical examples. OCIS codes: 310.0310, 310.1620, 310.6860, 230.4170, 230.4040, 350.2460.

In this paper, we report on the design, fabrication, and characterization of blazed gratings made tunable by electrostatic actuation. We combine KOH etching and deep-reactive ion etching (DRIE) to fabricate tunable blazed gratings (TBGs) on silicon-on-insulator wafers with optical-quality mirror surfaces on {111} crystalline planes. The actuators that are used to position the individual grating elements are designed to reduce rotation to levels acceptable for spectroscopic applications. We characterize the fabricated devices mechanically, verify the rotation-reduction design, and demunstrate the functionality of TBGs as tunable filters. [1552] Index Terms--KOH etching, microelectromechanical systems (MEMS), optical filter, optical grating, optical microelectromechanical systems (MEMS).

A multitap negative and positive coefficient radio-frequency transversal filter is implemented by using a digital-micromirror-device spatial light modulator for weighting-factor control and a chirped fiber Bragg grating for time-delay control. The demonstrated architecture is reconfigurable, has high speed and low loss, and is robust through digital programmability for a wide variety of filtering algorithms. A design using an interleaver for differential detection realizes an ultrahigh bandwidth with a maximum processable frequency of 33.7 GHz. A multitap low-pass filter, a negative tap notch filter with 40 dB attenuation, and a multitap negative coefficient bandpass filter are experimentally demonstrated. The results are in agreement with theory.

A new basic structure of a dispersive-compensation filter, called a chirped-cavity dispersion-compensator (CCDC) filter, was designed to offer the advantages of small ripples in both reflectance and group-delay dispersion (GDD). This filter provides a high dispersion compensation, like the Gires-Tournois interferometer (GTI) filter, and a wide working bandwidth, like the chirped mirror (CM). The structure of the CCDC is a cavity-type Fabry-Perot filter with a spacer layer (2 mH or 2 mL) and a chirped high reflector. The CCDC filter can provide a negative GDD of -50 [fs.sup.2] over a bandwidth of 56 THz with half the optical thickness of the CM or the GTI. OCIS codes: 320.0320, 310.0310.

Methods for the manufacture of rugate filters by the ion-beam-sputtering process are presented. The first approach gives an example of a digitized version of a continuous-layer notch filter. This method allows the comparison of the basic theory of interference coatings containing thin layers with practical results. For the other methods, a movable zone target is employed to fabricate graded and gradual rugate filters. The examples demonstrate the potential of broadband optical monitoring in conjunction with the ionbeam-sputtering process. First-characterization results indicate that these types of filter may exhibit higher laser-induced damage-threshold values than those of classical filters. OCIS codes: 310.0310, 310.1860, 350.2460.

A method to reduce the optical degrading effect of coating stress in a thin-film filter by releasing the filter from the substrate on which it was deposited is described. An extremely challenging 8-skip-0 100 GHz wavelength-division-multiplexing band-splitting filter is demonstrated. America OCIS codes: 310.1860, 310.6860.

The monitor passband width of an optical monitor is an important parameter for the fabrication of a dense-wavelength-division-multiplexing (DWDM) filter. The peak insertion loss and transmittance of one-cavity narrow-bandpass filters (NBPFs) were analyzed by using different passband widths. The simulation monitoring curves of the last layer of the first, second, third, and fourth cavities of a five-cavity DWDM filter with different monitor passband widths were investigated. The last layer of each cavity is very sensitive to the monitor passband width, showing that the monitor passband width of an optical monitor should be less than half the designed passband width. This analysis demonstrates the successful fabrication of a five-cavity DWDM filter. OCIS codes: 310.0310, 310.01860, 060.4510.

Vacuum-ultraviolet (VUV) narrow-bandpass filters with central wavelengths at 177.5, 180.7, and 193 nm are necessary for analytical chemistry applications and for atomic emission spectrum separation. The emission spectra are so close that the required passbands of these filters are i to 2 nm with a maximum of 2 nm. We first investigated the designs based on the Fabry-Perot model with fluoride materials and second the deposition of narrow-bandpass filter components by using evaporation deposition. VUV spectral characterizations show that the filter components meet the requirements for analytical chemistry applications. OCIS codes: 310.1620, 310.1860, 310.6860.

Solid-spaced filters are composed of one or several thin wafers of excellent optical quality acting as Fabry-Perot spacer layers. We study the different steps of the design and the manufacture of filters following dense-wavelength-division-multiplexing specifications. The design method of such filters requires a tight synergy between numerical simulations and experimental characterizations to correct possible thickness errors. Experimental results of the manufacture and characterization of a three-cavity narrow-bandpass filter and of an interleaver filter are given. OCIS codes: 060.4510, 060.1810, 230.5750, 350.2460, 310.6860.

The thermal shift of dense-wavelength-division-multiplexing (DWDM) filters that is caused by nonconstant temperature during fabrication invokes a spectrum shift and an error in the optical thickness of the layer before the matching layer. The relation between the thermal shift and the optical thickness of the layer before the matching layer was calculated. Five kinds of narrow-bandpass filters were fabricated, and their thermal shifts were measured and compared with the theoretical model. By using this technique, we were able to predict the thermal shift of DWDM filters from the abnormal behavior of the layer before the matching layer during deposition. OCIS codes: 310.0310, 310.1860, 060.4510.

The theoretical study and the experimental realization of an ultranarrow bandpass filter, joining a fiber Bragg grating and a dielectric mirror directly deposited at the extremity of the fiber tip, is presented. The features of such a filter are in very good accordance with the results of theoretical simulations. OCIS codes: 050.2230, 050.7330, 310.1620, 310.6860.

The ambiguity function was employed as a merit function to design an optical system with a high depth of focus. The ambiguity function with the desired enlarged-depth-of-focus characteristics was obtained by using a properly designed joint filter to modify the ambiguity function of the original pupil in the phase-space domain. From the viewpoint of the filter theory, we roughly propose that the constraints of the spatial filters that are used to enlarge the focal depth must be satisfied. These constraints coincide with those that appeared in the previous literature on this topic. Following our design procedure, several sets of apodizers were synthesized, and their performances in the defocused imagery were compared with each other and with other previous designs. OCIS codes: 110.0110, 110.1220, 110.4850.

A resonant-cavity-enhanced photodiode with broad filter transmittance and high quantum efficiency was numerically designed and analyzed, fabricated, and validated experimentally. We show theoretically that the quantum-efficiency spectrum broadens because of anomalous dispersion of the reflection phase of a mirror in the device and describe conditions that allow maximal flatness of the transmitted spectrum to be achieved. To demonstrate the concepts we design, fabricate, and characterize a backilluminated [In.sub.0.47] [Ga.sub.0.53] As-based p-i-n photodiode upon a InP substrate. Experimental measurements of the fabricated devices demonstrate a peak quantum efficiency of 0.80 at 1550 nm and a FWHM of transmittance of 35.96 nm. OCIS codes: 040.5160, 230.5160, 230.5170, 230.5750, 310.1620.

We report an experimental study of porous silicon-based rugate filters. We performed filter apodization, following a half-apodization approach, which successfully attenuated the sidelobes at both sides of the photonic stop band. We achieved successful reduction of interference ripples through the insertion of index-matching layers on the first and last interfaces. An apodized dielectric mirror and a rugate filter are compared: Appreciable differences in the harmonic presence and stop-band performance were observed and are commented on. Bandwidth control when index contrast is modified is also demonstrated. Finally, the possibility of combining different rugate filter designs to attain more complex responses is demonstrated by the achievement of a multi-stop-band filter. Numerical calculations for design optimization and comparison with experimental data are reported too. OCIS codes: 220.4610, 160.6000, 350.2460, 220.1230, 230.4170.

We have developed a wideband tunable optical filter that uses a long-period fiber grating (LPFG) in which both resonance wavelength and its signal attenuation can be adjusted. We create the grating mechanically by pressing a spring coil to an optical fiber. We achieve continuous fine tuning of wavelength and attenuation by varying the temperature of the LPFG. The adjustable ranges of the LPFG are more than 200 nm in resonance wavelength and more than 10 dB in signal attenuation. OCIS codes: 050.2770, 060.2310, 060.2340, 120.2440.

Details for the marginalized particle filter for a general nonlinear state-space model are discussed. It is demonstrated that the complete high-dimensional system can be based on a particle filter using marginalization for all but three states.

The dispersive characteristic of an electronically tunable optical filter is presented with application toward delaying high-speed digital and analog signals. It is shown that for 2-GHz square waves, dispersion is not a limiting factor and the filter can be operated over the full range of delays.

We describe a nonpolarizing filter design at oblique incidence and a polarizing filter design at normal incidence that use a uniaxially anisotropic layer. The phase thicknesses and the optical admittances of the layers are compensated for by the birefringent properties of a thin film at oblique incidence. This concept can be applied to the design of nonpolarizing bandpass and edge filters at oblique incidence and of polarizing beam splitters at normal incidence. Besides, the dependence of narrow-bandpass filters on normal incidence is discussed.

We discuss a simple method for fabricating interference birefringent filters using common cellophane tape layers. Cellophane tape layers can be superimposed with different orientations to generate different spectral responses. We demonstrate this behavior with a portable spectrophotometer. This technique is a simple and inexpensive way of investigating the optical properties of birefringent filters.

The design, fabrication, and characterization of a one-dimensional photonic crystal optical filter that has a relatively narrow, flat-topped passband within a wide stop band and small angular sensitivity is presented. The filter is based on a one-dimensional photonic crystal structure that has multiple defects, facilitating simultaneous minimization of the angular sensitivity and optimization of the passband's characteristics. We use epitaxially grown and selectively oxidized GaAs/[Al.sub.x][O.sub.y] multilayers to achieve a high-index-contrast material system and incorporate the experimentally determined optical and material properties into the design of the device. A flat-topped bandpass filter with a bandwidth of 65 nm and a wide field of view of 50[degrees] is experimentally characterized and compared with the design predictions. OCIS codes: 310.6860, 350.2460, 310.1860, 230.4170.

4f pulse shapers have been widely used to temporally manipulate femtosecond optical pulses by spectral filtering. When the temporal waveform is manipulated with a spatial light modulator consisting of segmented pixels, the spatial profile of the output beam also varies because of diffraction at the pixel array, which is known as a spatiotemporal coupling effect. This effect produces a complicated spatiotemporal profile near the focus of the ultrashort pulses, which may affect the interpretation of experimental results obtained with shaped ultrashort pulses. We investigate the spatial intensity distribution at the focus of temporally shaped pulses through ablation experiments. The three-dimensional space-time beam profile is also numerically calculated. OCIS codes: 320.5540, 220.4000.

An improved method is proposed to perform contouring of an object based on phase-measuring profilometry with a grid grating. Two phases unwrapped in perpendicular directions are obtained with the help of an adaptive bandpass filter and are used to unwrap the phases and to appoint the edge points of a shaded area and the object area. The height distribution of the object is obtained with the geometric relationship between coordinates and phases. The main axis of the projector and the main axis of the camera are not crossed and are also not in the same plane in order to arrange a measuring system easily and conveniently. The experimental results show that this technique is available for practical applications. OCIS codes: 070.2580, 100.2000, 120.5060.

A design of finite-size linear-phase separable kernels for differentiation of discrete multidimensional signals is presented. A numerical procedure for optimizing a set of principled constraints for multidimensional filter design is developed, and its use in the design of a set of high quality filters is demonstrated.

A study of an advanced new filtering scheme for processing the chromatic content of color images, encoded by the CIE u'v' chromaticity coordinates is discussed. The characteristics of the new filtering scheme are also described, and experimental results show that it is a valuable tool for smoothing and regularizing chromatic signals bringing forth a new approach to color image filtering.

We introduce what is to our knowledge a new nonlinear shift-invariant classifier called the polynomial distance classifier correlation filter (PDCCF). The underlying theory extends the original linear distance classifier correlation filter [Appl. Opt. 35, 3127 (1996)] to include nonlinear functions of the input pattern. This new filter provides a framework (for combining different classification filters) that takes advantage of the individual filter strengths. In this new filter design, all filters are optimized jointly. We demonstrate the advantage of the new PDCCF method using simulated and real multi-class synthetic aperture radar images. OCIS codes: 100.6740, 100.4550, 100.5010.

A reliable technology for optical resonant filters can be useful for optoelectronics technologies, optical communications systems, and biomedical applications. Positioning the resonance at a specific target wavelength is challenging because of the sensitivity of placing a narrow spectral filtering window to the error factors that are associated with the fabrication and characterization processes. We describe and prove experimentally two fabrication approaches to overcome this challenge. OCIS codes: 120.2440, 220.4610, 230.3120, 230.7370.

Images taken in noncooperative environments do not always have targets under the same illumination conditions. There is a need for methods to detect targets independently of the illumination. We propose a technique that yields correlation peaks that are invariant under a linear intensity transformation of object intensity. The new locally adaptive contrast-invariant filter accomplishes this by combining three correlations in a nonlinear way. This method is not only intensity invariant but also has good discrimination and resistance to noise. We present simulation results for various intensity transformations with and without random and correlated noise. When the noise is high enough to threaten errors, the method trades off intensity invariance in order to achieve the optimum signal to noise ratio, and the peak to sidelobe ratio in the presence of clutter is always greater than one. In the presence of random disjoint noise, the signal to noise ratio is independent of the target contrast and of the level of the noise. OCIS codes: 070.0070, 070.4550, 100.2000, 100.5010, 150.2950.

For a given low-pass analysis filter in an n-dimensional (n-D) perfect reconstruction (PR) FIR filter bank, there exists certain degree of freedom in designing a synthesis filter to make the overall filter bank have PR property, and Quillen-Suslin Theorem can be used to precisely determine the degree of freedom. Designing a PR synthesis filter with optimal frequency response amounts to minimizing the deviation between the frequency responses of the synthesis filter and an ideal filter under the PR constraint. In this paper, a method based on Grobner bases computation is developed for optimally designing critically sampled n-D PR FIR filter banks when one analysis filter is known. This design method can be easily combined with various existing design methods including minimax optimization and weighted least squares optimization, and can incorporate additional design goals including linear phase and bandpass characteristic. Index Terms--Grobner bases, n-D FIR filters, optimal frequency response, perfect reconstruction, Quillen-Suslin Theorem.

Large group delay in optical filters is examined using Fiber Bragg gratings as examples.

This study investigated the effects of using coloured filters on reading speed, accuracy, and comprehension as well as on perception of academic ability. A double-masked, placebo-controlled crossover design was used, with subjects being assessed over a period of 20 mo. There were three treatment groups (Placebo filters, Blue filters, and Optimal filters) involving 113 subjects with 'reading difficulties', ranging in age 9.2 yr. to 13.1 yr. and with an average discrepancy between chronological age and reading age of 1.8 yr. The 35 controls (who did not use coloured filters) ranged in age from 9.4 yr. to 12.9 yr., with an average discrepancy between chronological age and reading age of 2.1 yr. The treatment groups increased at a significantly greater rate than the control group in reading accuracy and reading comprehension but not for speed of reading. For self-reported perception of academic ability, two the the three treatment groups showed significantly greater increases than the control group. The larger improvements for treatment groups in reading comprehension may be related to a reduction in print and background distortions allowing attention to be directed to the processing of continuous text rather than to the identification of individual words. A reduction in print distortion, however, may not be sufficient to generate improved work-identification skills without additional remedial support, and this may be indicated by the nonsignificant increase in rate of reading.

Porous silicon Fabry-Perot optical filters with reflectivity peaks in the visible and near infrared spectral range were produced electrochemically and characterized with variable angle spectroscopic ellipsometry. Generalized ellipsometry and the Bruggeman effective medium approximation were used to analyze the anisotropy of the samples including determination of the tilt of the optical axis relative to the sample normal. At a wavelength of 700 nm, the difference between the real parts of the ordinary and extraordinary indexes of refraction of the low and high index sublayers were 0.05 and 0.07, respectively.

Transmission and phase-shift characteristics of dichroic high-pass filters with cutoff frequencies as high as 1.11 THz and of a cross-shaped grid bandpass filter with a resonance frequency of 280 GHz were measured with an electro-optic sampling terahertz time-domain spectrometer operating between 0.1 and 2 THz. Good agreement with transmission theories is found. We also compare the transmission performance of cascaded dichroic filters with that of cross-shaped grid bandpass filters. Both types of bandpass filter permit frequency-selective ultrafast experiments in the far-infrared spectral region. In the millimeter and the submillimeter wavelength regions, which are difficult to access by conventional means, knowledge of the frequency response of frequency-selective components is important for applications in frequency mixing, multiplying, and multiplexing in quasi-optical systems.

A dual-wavelength optical head, believed to be new in design, for 0.6- and 1.2-ram-thick-substrate optical disks was developed by use of 650- and 780-nm-wavelength light and a wavelength-selective filter, which reduces the spherical aberration that is due to the difference in substrate thicknesses and restricts the numerical aperture for 780-nm-wavelength light. According to this configuration, both high light power efficiency and wide image field characteristics are obtained, which are suitable not only for read-only but also for recordable or rewritable optical disk systems. A reading operation for a digital video disc and for a compact disc and a recording operation for a phase change optical disk were successfully demonstrated.

A new class of superresolution pairs of pupil filters for three-dimensional, two-pupil confocal imaging is proposed. A distinctive feature of these filters is the asymmetry of their impulse response. For synthesizing the amplitude transmittance of such filters the Fourier transform properties of Hermitian functions are employed. It is shown that, with simple phase-only filters that belong to the class in question, either axial or unidirectional lateral superresolution is achieved.

A polarization-based tunable interferometric filter essentially consisting of a two-beam interferometer with birefringence elements is described. The analysis of the filter is done through the concept of a geometric phase in optics - namely, the Pancharatnam phase. The transmission characteristics of the filter can be controlled through three parameters: the thickness of the birefringent elements, the optical path difference, and the orientation angle of an analyzer placed at the interferometer output. It is demonstrated theoretically that, with a particular choice of these parameters, the chromatic dispersion of the filter is compensated in a given spectral range. Some properties of the device are confirmed by an experimental demonstration.

Generalized spin filters, including several directional filters such as the directional median filter and the directional binary filter, are proposed for removal of the noise of fringe patterns and the extraction of fringe skeletons with the help of fringe-orientation maps (FOM's). The generalized spin filters can filter off noise on fringe patterns and binary fringe patterns efficiently, without distortion of fringe features. A quadrantal angle filter is developed to filter off the FOM. With these new filters, the derivative-sign binary image (DSBI) method for extraction of fringe skeletons is improved considerably. The improved DSBI method can extract high-density skeletons as well as common density skeletons.

In many image-processing applications the noise that corrupts the images is signal dependent, the most widely encountered types being multiplicative, Poisson, film-grain, and speckle noise. Their common feature is that the power of the noise is related to the brightness of the corrupted pixel. This results in brighter areas appearing to be noisier than darker areas. We propose a new adaptive-neighborhood approach to filtering images corrupted by signal-dependent noise. Instead of using fixed-size, fixed-shape neighborhoods, statistics of the noise and the signal are computed within variable-size, variable-shape neighborhoods that are grown for every pixel to contain only pixels that belong to the same object. Results of adaptive-neighborhood filtering are compared with those given by two local-statistics-based filters (the refined Lee filter and the noise-updating repeated Wiener filter), both in terms of subjective and objective measures. The adaptive-neighborhood approach provides better noise suppression as indicated by lower mean-squared errors as well as better retention of edge sharpness than the other approaches considered.

The design of reduced-resolution filters by multiresolution analysis (MA) or by downsampling is extended to spatial light modulators with fill factors less than one. Analysis shows that the dependence of the zero-order correlation peaks with fill factor varies with target size for both design techniques. Also, a practicable performance improvement is obtained for MA compared with downsampling for small- to medium-sized targets that is greater for larger fill factors. The validity of the analysis is confirmed by simulation. A reduced-resolution optical correlator is constructed, and a comparison of MA and downsampling filters is performed for different-sized targets. The experimental results show good qualitative agreement with the simulation; however, the first-order correlation peaks were found to be greater for the experimental results. A possible reason for this is that the manufacturer's fill-factor specification might be too large; therefore a new technique for measuring the fill factor is proposed. OCIS codes: 100.1160, 100.4550, 100.1390, 100.5010, 230.6120, 100.7410.