Employment of optical techniques in signal processing and communication and computing systems has become a major research and development effort at many industrial, government, and university laboratories across the nation and in Europe and Japan. implementation of optical computing concepts and the use of bistable etalons and non-linear logic devices in computing have gained a lot of support and enthusiasm from the optics community in recent years. The significance Iof this field and its potential importance in future technologies is evidenced by the large number of conferences, workshops, and special issues on the subject.
Optics is under serious consideration with regard to the implementation of parallel multiprocessing systems, especially those involving a high degree of interprocessor interconnection. This paper discusses multiprocessing architectures and describes the developing technology directed toward advanced multiprocessor machines that is supported by the Defense Advanced Research Projects Agency (DARPA). The emerging device and materials technology is discussed. The major device development is categorized into four areas: sources and detectors integrated with electronics, two-dimensional spatial light modulators, optical beam steering, and bistable optical devices.
Shift invariance in the context of associative memories is discussed. Two optical systems that exhibit shift invariance are described in detail with attention given to the analysis of storage capacities. It is shown that full shift invariance cannot be achieved with systems that employ only linear interconnections to store the associations.
Boltzmann machines can be used to obtain solutions to problems for which direct mathematical calculation of a solution is not possible or would take too long. The range of problems for which such a machine might be useful is very broad, and inversion of a linear equation is described as one example. A method is described that allows time-intensive operations on a computer to be replaced with a highly parallel optical system that can be incorporated in such a way as to let the computer take full advantage of the speed of the optics without sacrificing the precision possible in the computer. A tie is made between Boltzmann machines and neural networks, and it is shown how a Boltzmann machine may perform functions demonstrated by other neural network architectures.
A bimodal optical computer (BOC) for solving a system of linear equations is presented. The BOC can achieve accuracies comparable to those of the digital computer, and its speed is far superior in solving a system of linear equations. The advantage in speed increases with the size of the matrix. The problem of the convergence of the solution using the BOC is investigated. It is found that by using a BOC with an error as high as 50% in the matrix's optical mask and 1% in the electro-optical devices, convergence is achieved for matrices with condition numbers of 25. The effect of the condition number on the convergence of the solution is investigated. It is found that matrices with large condition numbers converge very slowly. Convergence for matrices with condition numbers higher than 250 was achieved. A means of improving the condition number of a matrix is also introduced.
Spatial filtering is one of the main assets of optics for information processing. In this paper we review spatial filtering methods for performing binary logic operations. Many pairs of bits can be processed simultaneously. The input data are arranged in matrix form. The type of operation is usually homogeneous across the matrix. The input is characterized as a diffracting, as a scattering, or as a birefringent structure. Experimental results are shown. Applications of the described method for an optical adder are presented.
Definitions of symbolic processing, explicit and implicit declarative knowledge, and procedural knowledge are given. Architectures for optical symbolic correlation processors for pattern recognition problems are given, and the uses of explicit and implicit declarative knowledge as well as procedural knowledge are discussed.
Pattern recognition is demonstrated using two-dimensional nonlinear optical logic gate arrays on ZnS and ZnSe interference filters defined by fly's eye lens arrays. Fan-out is demonstrated, allowing symbol scription using these arrays.
We show experimentally how nonlinear all-optical circuit elements can be used to create indefinitely extensible optical logic. By using a "lock and clock" architecture and an off-axis configuration of the power and signal beams, we avoid signal corruption and maintain an undiminished signal level through cascaded devices. Several loop circuits have been operated to simulate an optical classical finite state machine.
It is essential to pixellate arrays of optically bistable switching elements in order to remove cross-talk caused by diffusion. We solve for the problem of cylindrical pixel nonlinear Fabry-Perot cavities, illuminated by nonuniform radiation beams, accounting for carrier diffusion and surface recombination. It is found that the additional carrier sink produced by the transverse surface is not detrimental to the switching power if the pixel area is small enough or if the surface recombination velocity is low enough.
This paper presents a new pseudocolor method that uses a special sensitive plate with a code. It is quite convenient to use the plate to convert a black-and-white picture into a density pseudocolor-coded image. Users need only place the black-and-white picture and the plate with a code in immediate contact with each other and expose them simultaneously. Through simple development, an encoded phase transparency is obtained. That transparency is put on the input plane of a white-light processor; the density pseudocolor-coded image on the output plane is produced by the spatial filtering.
This paper deals with an algorithm-driven architecture devoted to fast edge detection. The architecture has been specifically designed to process large convolution masks in a pyramidal (multiresolution) scheme. The basic element of the convolution board is a programmable VLSI component. Several identical components can be connected in a virtually systolic structure in order to achieve the desired throughput rate. A distinctive feature of the system is the multiple-resolution capability of the convolver board. The number of convolver boards hosted by a multiple bus vision machine can be selected to achieve a parallel multiple-resolution operation. The main application of the proposed architecture is for fast edge detection based on the extraction of the zero crossings of Gaussian filtered images. The paper is divided into two sections: the first presents results of numerical simulations, showing the accuracy of this edge-detection technique applied to convolved images in a pyramidal structure; the second presents the systolic architecture implementing the algorithm.
For staring sensors, improved performance in the location of point sources can be achieved by use of an array of hexagonal detectors instead of the usual array of square or rectangular detectors. This improvement is demonstrated by calculating the accuracy of the centroid algorithm as a function of signal-to-noise ratio and blur spot size for both types of detector arrays. The probability density function for the centroid random variable is derived and is used to perform all noise analysis. The analysis indicates that the algorithm error is reduced by as much as a factor of 3, the sensitivity to noise is reduced by 17%, the computational load is decreased by 23%, and the data storage requirement is reduced by 22%. The clutter-induced noise, as measured by the clutter equivalent target, is essentially identical for square and hexagonal detectors of the same area.
A new and versatile experimental stress analysis technique has recently been developed, based on the measurement of the infrared radiation emitted from the surface of a body as a result of temperature changes caused by cyclic loading within the elastic range ( i.e., the thermoelastic effect). The theory is outlined and the apparatus [referred to as the SPATE (stress pattern analysis by thermal emission) equipment] is described. A number of applications are described to illustrate the principal characteristics of the technique, which is now regarded as a valuable new approach in the stress analysis and design assessment of a wide range of engineering components and structures. Possible future developments in this area are outlined.
Lack of precision and accuracy of in-process critical dimension (CD) measurements of linewidth continues to be a serious problem at micrometer and submicrometer dimensions. Even with highly repeatable optical linewidth measurement systems, variable "offsets" or errors have been shown to occur with changes in process variables such as thickness of the patterned layer and sublayers and changes in the indices of refraction of the materials. All of these variations result in a change in the optical phase difference that occurs on reflection at the line edge and, therefore, result in changes in the structure of the optical image. Although an accurate coherent optical edge-detection method has been developed, it requires accurate knowledge of this phase difference, which is not always possible in CD measurements. This paper proposes a new dual-threshold method for edge detection and focusing, based on image theory, which can be adapted to most optical microscope based measurement systems. It does not require knowledge of the phase discontinuity at the line edge. The accuracy of this criterion is compared to two more widely used criteria, (1) the minimum and (2) 50% threshold, and it is concluded that, when the phase difference is unknown and varies with normal processing, the new dual-threshold method is the superior method.
As can be seen from the editorial schedule for Optical Engineering also found on this page, the special issues forth-coming in 1987 will address a variety of important and timely topics. Following a tradition established over the past several years, the first issue of the year is devoted to the general subject of optical computing. In his guest editorial on the facing page, Nasser Peyghambarian reviews the contents of the nine papers included.
Video Microscopy is addressed to the reader already somewhat experienced in light microscopy but perhaps with less of a background in video or electronics. The author, Shinya Inour, describes the organization of the book as a spiral, "designed to guide the reader through a web of practice and theory, in much the way I believe many of us learn in a new situation." he has accomplished his introductory objective in a book that is understandable and convienent to use.