This PDF file contains the editorial “Upon Proceeding...” for OE Vol. 37 Issue 06

Binary granulometric filters are formed from unions of parameterized openings, a point passing the filter if and only if a translate of at least one structuring element fits in the image and contains the point. A granulometry induces a reconstructive granulometry by passing any image component not eliminated by the granulometry. As historically studied in the context of Matheron's granulometric theory, reconstructive granulometries appear as unions of reconstructive parameterized openings. The theory is extended to a much wider class of filters: a logical structural filter (LSF) is formed as a union of intersections of both reconstructive and complementary reconstructive openings. A reconstructive opening passes a component if and only if at least one translate of the structuring element fits inside; a complementary reconstructive opening passes a component if and only if no translate of the structuring element fits inside. The original reconstructive granulometries form the special class of disjunctive LSFs. Complement-free LSFsform granulometries in a slightly more general sense; LSFs containing complements are not increasing and therefore not openings. Along with the relevant algebraic representations for LSFs, the theories of optimal and adaptive granulometric filters are extended to LSFs, a systematic formulation of adaptive transitions is given, transition probabilities for adaptation are found, and two applications to biological imaging are presented.

A new side-projected catheter for in vivo blood flow measurements has been developed and tested. This catheter is designed to measure blood flow in both the forward (towards the catheter tip) and reverse (away from the catheter tip) flow directions. It consists of two multimode optical fibres with core diameter of 50 xm and cladding diameter of 125 µm. One fibre transmits the laser beam into blood and the other receives the back-scattered light from the erythrocytes within the probe volume. The experimental plots of the dominant frequency (frequency with the highest amplitude) against the flow velocity are found to be linear in both the forward and reverse flow directions. This result was reaffirmed by the numerical flow simulation along the fibre catheter side wall. In the simulation, it was found that the flow within the boundary layer is indeed laminar, and hence the relationship between the Doppler shift frequencies and the flow velocities is linear, thereby making the linear calibration possible for predicting the free stream flow velocity. The analysis of the experimental and the computational results indicated that the boundary layer along the side wall of the fibre catheter varies with both the flow velocity and flow direction. Furthermore, using the fibre optic catheter with an increasing fibre core separation, it was observed experimentally and numerically that the percentage of the free stream flow velocity increases, suggesting that the projection of the measuring probe volume is indeed moving towards the free stream domain.

The vertical variation of extinction and backscatter in the lowest part of the atmospheric mixed layer is of considerable interest for the assessment of the performance of electro-optical systems, but is poorly known. Our purpose is to estimate the vertical variation of extinction and atmospheric transmission due to aerosol particles near the air-sea interface. Atmospheric extinction coefficients in the 3- to 5- and 8- to 12-µm IR bands calculated using vertical aerosol distributions measured in the Mediterranean coastal zone are presented. The data reveal a negative extinction gradient between 1 and 2 to 3 m height during winds of marine origin lower than 9 m/s, whereas a positive gradient occurs between 1 and 2 to 3 m height during high-wind-speed periods (V?9 m/s). A positive extinction gradient is also noted for vertical profiles recorded during moderate continental winds. These extinction gradients result in a relative difference in atmospheric transmission between two sample heights that reach 60% in both the 3- to 5- and 8- to 12-µm bands for an optical path of 10 km during high-wind-speed periods. These observed differences cannot be explained by current models to predict the vertical aerosol distribution.

Diffraction efficiencies of binary dielectric multilayer gratings are calculated by combining thin film theory and scalar electromagnetic theory, and results are compared with efficiencies of conventional binary gratings, consisting of one single bulk material. By optimizing the number of layers and layer parameters such as material and layer thickness, multilayer grating designs are obtained showing diffraction efficiencies in the desired order close to the theoretical maximum and suppressing the zeroth diffraction order effectively. To meet technological requirements of etching processes such as a suitable etch-depth control, parameters of etch-stop layers are included into the calculation and optimization procedure. Manufacturing dielectric multilayer gratings including etch-stop layers will make it possible to obtain uniform etch profiles, a moderate surface roughness in the etched area, and high accuracies in etch depth without any online control devices.

The concepts of multiresolution sampling (MRS) are explained, and relevant models are built. The Shannon models are suitable for processing signals with good mathematical properties, but our models are superior in dealing with complex signals, while retaining the capabilities of Shannon models. The main differences between the Shannon models and ours are different assumptions about sampling. The models can be well applied to the resampling of the digital images containing complex information, and the concepts can be brought into the scope of wavelet theory.

Bacteriorhodopsin-based holographic interferometry is a new technique with which high-resolution interferograms can be recorded at one or more wavelengths in real time. The erasable nature of bacteriorhodopsin (bR), its panchromatic response to light, and the fact that recording is done on the molecular level allow for the formation of fringe patterns combining the high resolution of conventional silver halide recording materials with the real-time properties of CCD cameras. In this context, we have used bR to image changes in the crystal environment at single and dual wavelengths, using a real-time sequencing architecture whereby successive exposures of a bR thin film are overlapped to produce a continuous stream of interferometric images. This allows realtime data buffering and immediate, on-line observation of results. In this paper, we present real-time single and dual wavelength interferograms of growing KAl(SO4)2 and melting sugar crystals recorded on bR thin films. The interferograms were produced with a shuttered cw argon or a pulsed Nd:YAG write laser, and a helium-neon read laser, and show contrast more than adequate for the given application. The results of this effort clearly demonstrate the real-time recording capabilities of bR thin films.

The quantum efficiency value is determined by studying the surface potential development in chalcogenide vitreous semiconductors. Sensitization by a corona discharge in response to exposure to radiation from a Q-switched nanosecond-pulse ruby laser is achieved. A method to calculate the quantum efficiency is presented, and its dependence on parameters such as the exposure moment switch, temperature, charging electrode potential, and radiation energy density is studied. The mobility of the holes is determined by the time of flight of carrier packets. The observed effects are explained using the residual conduction mechanism.

Fractal image compression is based on the self-similarity search of the image. The encoding process is computationally intensive. We present a fast fractal image encoding algorithm that is based on a refinement of the fractal code from an initial coarse level of a pyramid. Assuming that the distribution of the matching error is described by an independent, identically distributed (i.i.d.) Laplacian random process, we derive the threshold sequence for the objective function in each pyramidal level. The algorithm is quasi-optimal in terms of minimizing the mean square error. Computational efficiency depends on the depth of the pyramid and the search step size, and could be improved by up to two orders of magnitude over the computational effort required for a full search of the original image.

An ordered gray-scale erosion is suggested according to the definition of hit-miss transform. Instead of using three operations, two images, and two structuring elements, the developed operation requires only one operation and one structuring element, but with three grayscale levels. Therefore, a union of the ordered gray-scale erosions with different structuring elements can constitute a simple image algebra to program any combined image processing function. An optical parallel ordered gray-scale erosion processor is developed based on the incoherent correlation in a single channel. Experimental results are also given for an edge detection and a pattern recognition.

We propose a method for online measurement of section sizes of hot-rolled sections. We consider two main points. First, section sizes of the measured workpiece depend on its section outline; here the section outline is presented by perpendicularly intersecting the workpiece with an optical plane. Second, online measurement of section sizes is almost a real-time process; here to shorten the measurement period, the recording process and the initial extraction process of one section outline are combined by recording it as three 2-D binary patterns. These patterns come from three areas of the section outline and are sufficient for calculating the section sizes. The experimental performance of this method was tested and verified in our laboratory and in a rolling mill in China. A set of the tests are presented.

infrared thermal detectors were prepared with pyroelectric PSN-PT-PZ (1/47/52) ceramics, where a signal electrode had a structure Au/metallic buffer/Pb(Zr,Ti)O3 ceramic. The effect of a metallic buffer layer on the voltage responsivity was investigated with a response to a step signal, made by a dynamic pyroelectric measurement. A pyroelectric ceramic wafer was prepared by a mixed-oxide technique. The Au layer (thickness 50 nm) and the metallic buffers (thickness 0 to 20 nm) of Cr, NiCr (80 : 20), and Ti were prepared by dc magnetron sputtering. In order to improve the light absorptivity, Au black was coated on the Au signal electrode by thermal evaporation. A detector without a buffer layer showed a noisy and fluctuating output signal. Among the three kinds of buffer materials, NiCr (80 : 20) and Ti adhered well with ceramics and showed good electrical and thermal contacts, whereas Cr resulted in bad contacts. Considering the output voltage and thermal properties, the optimum thickness of the buffer layer was about 15 to 20 nm, and sensors with a Ti buffer 15 to 20 nm in thickness showed good detectivity. Thus, the stability and reliability of the infrared thermal sensors could be improved by using an appropriate buffer layer.

A theoretical analysis of the Auger mechanism in a GaInAsSb infrared photovoltaic detector is reported. The considerations are carried out for near-room temperature and 2.5-µm wavelength. The results show that the Auger mechanism can be suppressed by optimizing the material parameters. Thus, the performance of such detectors can be improved.

It is a well-known phenomenon that three- and four-bar patterns can sometimes be resolved with staring sensors even when the bar pattern frequency is beyond the half sample rate of the sensor. When performing a minimum resolvable temperature (MRT) test, for example, the modulation of the target bars can be significant even when the fundamental frequency of the four-bar target is higher than the half sample rate of the sensor. We show that the modulation of a four-bar target goes to zero at 0.6 times the sample rate of the sensor (1.2 times the half sample rate) if the sample phasing is optimized. If the bar pattern contains a larger number of bars, the modulation cutoff approaches the half sample rate of the sensor. Further, we illustrate that, when the sample phase is optimized during the MRT measurement, the MRT performance of a sampled imager can exceed the measured MRT performance of an analog sensor with the same system and component modulation transfer functions.

A novel compact ultrafast tunable fiber-loop laser is proposed, with wavelength tuning speeds of a few nanoseconds. The laser implementation is based on currently available technologies. With the use of compound configurations, the laser tuning range can be extended to over 50 nm, and a sidemode rejection ratio of over 50 dB can be achieved.

A quantum theory of the spectral width of a laser with a saturable absorber is presented. We perform a density matrix calculation yielding the off-diagonal element pn,n+1(t ) which is proportional to the average value of the electric field (E(t )) associated with the laser field having a large photon number n. The general linewidth theory is applied to any general laser which has saturable absorber features for a portion of the building of the laser field. We have found that the saturable absorber action contribution to the linewidth of such a laser can be substantial.

In manufacturing processes, various and suitable pulse shapes are required for the purpose of materials processing, and the pulse shape is regarded as a dominant factor according to the specific property of materials. Therefore, this study has undertaken to generate the correct ones by using the two-pulse superposition technique. The pulsed Nd:YAG laser using a multi-mesh inductor L and capacitor C network and the pulse superposition technique is described. We have tried to superpose two pulses with the single shot multivibrator (SSM: 74LS123) having two silicon-controlled rectifiers (SCRs) which control the delay time of the two gate turn-on signals. One is a main rectangular pulse, and the other is a superposed sinusoidal pulse. The use of this technique might impart a number of advantages to people who need a suitable pulse shape for particular applications such as welding, cutting, and drilling. In addition, we compared the case of no superposition and applied input energy of 100 J in only the main circuit with that of superposition and applied input energy of 75 J in the main circuit and 25 J in the superposing circuit. It is found that the maximum efficiency of 4.5% is obtained within the range of delay time less than 230 µs by adjusting the delay time from 0 to 1 ms.

A new interferometric system for measuring straightness is developed. In the system, triangular prisms are used to generate two converging beams. A straightness reflector is used to reflect back the converging beams along their respective paths to form a combined beam. The combined beam is received by a detector system to count the interference fringes. Being different from the traditional straightness reflector, the new straightness reflector consists of not only reflecting surfaces but also refractive surfaces. The refractive surfaces cause a change in optical path difference (OPD) when a relative lateral displacement between the triangular prisms and the straightness reflector takes place. The relationship between the OPD and the lateral displacement is derived. The expected results agree with the results in the experiments. A straightness measurement with an accuracy of 0.1 µm is obtained over a travel range of 120 mm.

A method to measure the derivative of displacement using time variation changes in the object together with Fourier transform analysis in speckle shear interferometry is presented. The concept of the method is that the object is deformed continuously and a large number of sheared images of the object motion are acquired using a high speed CCD camera. The derivative of the object deformation is then retrieved from this large set of data using Fourier transformation. The method is capable of obtaining information for object displacements over 500 µm, which is a very difficult task when using conventional electronic speckle pattern shearing interferometry. Theory as well as some of the experimental results with the new method are delineated.

We describe an optical system for 3-D profilometry based on the white light interferometer. Recently many different methods have been used to analyze the data obtained from white light interferometric profilers. Many commercially available white light profilers are also in use today. We detail a simple way to construct a profiler that uses two simple and efficient algorithms. It deals with the data in a fast and simple manner, thus reducing both the acquisition and analysis time. The system has a theoretically unlimited range and can profile both optically rough and smooth surfaces.

An experiment is conducted on using laser mode splitting technology to measure displacement. A laser longitudinal mode is split into two frequencies (modes) by the birefringence effect of an intracavity quartz crystal wedge. The magnitude variation of frequency splitting is proportional to the displacement of the wedge crossing the intracavity laser beam. The dependence of sensitivity and measurement range on the wedge angle and angle between crystalline axis and intracavity laser beam is studied. A resolution of 10 nm is obtained.

A simple technique for the replication of micro-optical elements is presented. Elastomeric material is used to realize a negative mold of the original optical element and UV-curing adhesive is used to make the replicated copy. Replicated elements, such as refractive and diffractive micro-optical elements, are produced. The refractive microlenses have a diameter of 970 µm and a height of 79 µm and the diffractive element is a multilevel blazed grating with a period of 64 µm. The characteristics of the replicated elements are measured using different methods. The deviation from a sphere of the original and the replicated refractive microlenses is 0.13?. and 0.12?. root mean square (rms), respectively. The replicated multilevel blazed grating has a diffraction efficiency of over 80%. An alternative method for realizing planoconcave microlenses from a planoconvex master is also presented.

The optical analysis and design program RAYTRACE, which was developed at our institute, enables the simulation of micro-optical systems containing arrays of microlenses or other optical components. The most important features of the program are briefly described to show its broad application areas. A special simulation example showing the spot diagram and the intensity distribution in the focal plane of a simple Shack-Hartmann sensor with a diffractive microlens array for an incident aspherical wave is given.

An adaptive clipping algorithm with asymmetric clipping points is proposed. By introducing the Gram-Schmidt orthogonal projection algorithm into the Hopfield model and then using the asymmetric adaptive clipping method, we have constructed an asymmetric adaptively clipped model. statistical simulations have revealed that the asymmetric adaptively clipped model has relatively strong performance for patterns with unequally distributed neuron states and of great similarity. since there are only three values used in the interconnection weights for the asymmetric adaptively clipped model, the interconnections are much more easily implemented with the optical technique than in the original Hopfield model. A 2-D hybrid optical neural network with 1024 neurons, in which a beam-direction encoding method is employed, is proposed to implement the adaptively clipped model. Preliminary experimental results for the reconstruction of traffic signs are shown.

We investigate an optoelectronic system that can perform weighting and threshold operations similarly to a neuron. The weighted summation is carried out optically in an optical fiber by using stimulated Brillouin scattering. It is shown that magnitude and sign of a weight associated with an input can be set by changing the signal intensity and wavelength, respectively. A signal traveling in the backward direction is required to sense additions and subtractions. Due to inherent parallel nature of optical processing, this system can outperform a digital system for large numbers of inputs.

A method is proposed in which handwritten characters are recognized using feature extraction and an associative memory matrix. In feature extraction, simple processes such as shifting and superimposing patterns are executed. A memory matrix is generated with singular value decomposition and by modifying small singular values. The method is optically implemented with two liquid crystal displays. Experimental results for the recognition of 26 handwritten alphabet characters clearly show the effectiveness of the method.

In a many-to-many collimating system using gradient-index- rod lenses (GRINs) for an optomechanical matrix switch, there are multiple cross-connections having different optical path lengths so that excess coupling losses arise. The total coupling loss is greatly reduced by properly choosing the GRIN lengths. Two schemes can do so. In the equal ?Z scheme, all GRINs have the same excess length ?Z0 than the quarter pitch, which is optimized at 13, 25, and 49 µm for the three focusing parameters of 0.237, 0.326, and 0.417 mm-1, respectively. The total coupling loss is further reduced in the unequal ?Z scheme, in which the ?Zµ of each GRIN is independently determined and optimized by numerical computation. The results are preferred rather than the zero ?Z scheme where just quarter-pitch GRINs are used.

A 5-bit true-time delay module is demonstrated. The combination of a 1-to-4 fiber beamsplitter and four 1-to-8 collinear guided-wave fanouts provides us with the required 32 surface-normal fanout beams with a delay step of 50 ps. The system design, device fabrication, and optimization of the fanout intensity uniformity down to ±10% among 32 fanouts are addressed. The packing density of this device is 2.5 (delay lines)/cm2, which is the highest demonstrated thus far. A bandwidth of 2.4 THz is confirmed experimentally by employing a femtosecond laser and an ultrafast detector.

The new approach to phase-stepping photoelasticity known as ‘‘load stepping’’ is used to determine automatically the isochromatic parameter a and the isoclinic angle ?. There is no need for the user to calibrate the results other than to convert the isochromatic parameter into a principal stress difference using the material fringe constant. Four phase-stepped images are collected using a circular polariscope for each of three load steps, which differ by small equal increments. A ramped phase map for the isochromatic parameter is produced in the range -?