Abstract: Segmentation of an image into separate regions is useful in many computer vision problems. The separate regions may be represented by their enclosing contours. A planar graph is constructed for each image. The optimal contour is determined by the assignment of edge costs and face weights of the graph. An algorithm is provided such that given a planar graph G = (V,E) where V is a set of nodes, E is a set of edges and each edge e is provided with a non-negative cost c(e) and each face f of the planar graph is provided with a non-negative weight w(f), and given a single node v in G and single face F adjacent to v which is viewed as the exterior face, then find a directed path P in G that starts and finishes at v and minimizes ##EQU1## where cost(P) is defined as the length of the path under c, and weight (P) is the weight of the faces that are separated from F by P. The contour that minimizes .zeta. is the optimal contour.

Abstract: A watermark is embedded into video/image/multimedia data using spread spectrum methodology. The watermark is extracted from watermarked data without the use of an original or unwatermarked version of the data by using MPEG/JPEG coefficients. The image to be watermarked is divided into subimages. Each subimage is embedded with a watermark. When extracting the watermark, the result from each subimage is combined to determine the originally embedded watermark.

Abstract: A metallic film has apertures located therein in an array arranged in a pattern so that when light is incident on the apertures, surface plasmons on the metallic film are perturbed resulting in an enhanced transmission of the light emitted from individual apertures in the array. The aperture array is used: to filter light of predetermined wavelength traversing the apertures, to collect light over a distance after traversing the apertures, to improve operation of near-field scanning optical microscopes, and to enhance light transmission through masks useable in photolithography.

Abstract: A self-biasing, non-magnetic giant magnetoresistive sensor having a Corbino-disk geometry constructed from a thin film of e.g., doped, Mercury Cadmium Telluride (MCT) Hg.sub.1-x Cd.sub.x Te exhibiting anomalously large Giant Magnetresistance (GMR) and zero field offset. In one embodiment, the sensor has a silicon substrate, a layer of doped, inhomogeneous MCT, and electrodes attached to the inhomogeneous layer. Alternatively, a buffer layer of, e.g., CdTe may overlay the substrate. In another embodiment, the sensor has a silicon substrate, a layer of doped, homogeneous MCT, and electrodes attached to the doped homogeneous MCT. Alternatively, a buffer layer of, e.g., CdTe may overlay the substrate as well. With constructed in as either of these embodiments, highly doped Corbino devices may show a significant zero-field offset in the GMR which results in a built-in bias field as high as 1500 G at T=300 K.

Abstract: A method for solving the stereo correspondence problem transforms the problem into a maximum-flow problem. Once solved, the minimum-cut associated to the maximum-flow results in a disparity surface for the entire image at once. This global approach to stereo analysis provides a more accurate and coherent depth map than the traditional line-by-line stereo method. Moreover, the optimality of the depth surface is guaranteed and can be shown to be a generalization of the dynamic programming approach that is widely used in standard stereo. Results show improved depth estimation as well as better handling of depth discontinuities. While the worst case running time is O(n.sup.3 d.sup.3), the observed average running time is O(n d.sup.1.4) for an image size of n pixels and depth resolution d.

Abstract: A low-cost star-coupler assembly primarily for use in optical data networks is fabricated from three modular components. The first component is a support plate having at least two sets of grooves into which polymer optical fibers may be placed and a central aperture. The second component is a waveguide plate which is designed to fit in the aperture. The third component is a cover plate having alignment means for aligning the components in fixed relationship and for holding polymer optical fibers in the grooves securely in place. The coupling structure is most useful with large-core polymer optical fibers. The modular design facilitates replacement of fibers. The components may be fabricated by injection molding thereby reducing the cost.

Abstract: A subwavelength-resolution optical imaging device is provided for conveying light having a wavelength .lambda. that is emitted, reflected or transmitted by a sample. The device comprises a coherent fiber bundle, which itself comprises a plurality of optical fibers disposed substantially coaxially along their lengths. Each optical fiber also comprises a core having an index of refraction n.sub.1 and a cladding having an index of refraction n.sub.2 thereby providing a value .DELTA., where: ##EQU1## Light is effectively confined within the core of each fiber without need for a separate coating applied to the fibers by selecting the values of n.sub.1 and n.sub.2 for each optical fiber so as to provide a value for .DELTA.

Abstract: A multistage database search method includes a first stage that only considers textual polygraphic frequencies and a second stage that considers the location of the polygraphs and coarsely approximates the proper alignment between the query and the database strings and a third stage that provides a better approximation of the alignment. The final output is a listing of suggested database records ranked by their similarity to the query.

Abstract: A metallic film has apertures located therein in an array arranged in a pattern so that when light is incident on the apertures, surface plasmons on the metallic film are perturbed resulting in an enhanced transmission of the light emitted from individual apertures in the array. The aperture array is used: to filter light of predetermined wavelength traversing the apertures, to collect light over a distance after traversing the apertures, to improve operation of near-field scanning optical microscopes, and to enhance light transmission through masks useable in photolithography.

Abstract: The low-field magnetoresistance of a high carrier mobility semiconductor with inhomogeneities which are more conducting than the surrounding semiconductor material matrix is significantly enhanced compared to the magnetoresistance of the homogeneous material. The enhancement results from a magnetic field induced geometric effect in which current is expelled from the conducting inhomogeneity. The enhanced giant magnetoresistance is demonstrated at low field in (near) zero-band-gap material, such as Hg.sub.1-x Cd.sub.x Te(x.about.0.1). The effect is applied to the fabrication of magnetic read head sensors such as Corbino disc, bar magnetoresistance sensors and thin film sensors.

Abstract: A synchronous test model (STM) and corresponding method capture the essential behavior of an asynchronous circuit under test. During operation of the method, (1) An STM for the asynchronous circuit is constructed assuming either a user-specified cycle length or an estimated cycle length; (2) a target fault list is created containing only faults in the asynchronous circuit, (3) test patterns are generated from the STM using a synchronous test generator; (4) the test patterns are translated into test sequences for the asynchronous circuit; and (5) the translated patterns are validated by fault simulation on the asynchronous circuit.

Abstract: Based on the encoding of deterministic finite-state automata (DFA) in discrete-time, second-order recurrent neural networks, an algorithm constructs an augmented recurrent neural network that encodes a FFA and recognizes a given fuzzy regular language with arbitrary accuracy.

Abstract: Digital watermarking of audio, image, video or multimedia data is achieved by inserting the watermark into the perceptually significant components of a decomposition of the data in a manner so as to be visually imperceptible. In a preferred method, a frequency spectral image of the data, preferably a Fourier transform of the data, is obtained. A watermark is inserted into perceptually significant components of the frequency spectral image. The resultant watermarked spectral image is subjected to an inverse transform to produce watermarked data. The watermark is extracted from watermarked data by first comparing the watermarked data with the original data to obtain an extracted watermark. Then, the original watermark, original data and the extracted watermark are compared to generate a watermark which is analyzed for authenticity of the watermark.

Abstract: Digital watermarking of data, including image, video and audio data, is performed by repeatedly inserting the watermark into subregions or subimages of the data. Similarly, the watermark is repeatedly extracted from the subregions of the data.

Abstract: A watermark is embedded into audio/video/image/multimedia data using spread spectrum methodology. The watermark is extracted from watermarked data without the use of an original or unwatermarked version of the data by using spatial or temporal local averaging of the frequency coefficients of the watermarked data.

Abstract: A computer technique for bipartite matching of objects of one subset with objects of a different subset where multiple choices are permitted. A bipartite graph is formed in which the objects form nodes and the edges connecting pairs of nodes represent costs of matching the nodes connected. The original tour or graph is decomposed into a plurality of quasi-convex subtours or subgraphs and the minimum cost match of each subtour is found and the union of all such matches of the subtours is used as the desired match.

Abstract: Single- or double-strand oligonucleotides are used to create a molecular automata. The preferred embodiment is a DNA Turing machine and a method of performing a transition in such a DNA Turing machine.

Abstract: A heterojunction is formed between a pair of layers of different semiconductive materials whose work function difference produces a large band offset at the heterojunction. Donor or acceptor atoms are included in one regions that when photoexcited produce free charge carriers but leave behind charged centers that keep the photoexcited carriers localized. The large barrier at the heterojunction limits recombination of the free charge carriers and the charged centers and persistent photoconductivity results. This effect is used to form light operated switches. An illustrative example uses a layer of high purity gallium arsenide forming a heterojunction with a gallium-doped layer of zinc selenide.

Abstract: In a supercomputer that employs a large number of processors operating in parallel, groups of processors are assembled in clusters. Each processor in a cluster is provided with an array of lasers, a different one targeted for a particular detector associated with a specific processor of a different cluster. The detectors associated with the processors in the different cluster are assembled into an array of detectors. A processor wishing to communicate with a processor in a different cluster excites the laser in its associated array that is targeted to the detector of the processor with which it intends to communicate. The patterns of light of the various laser arrays are superimposed into one pattern focused on the detector array by means of an optical tunnel and a spherical lens.

Abstract: A tapered probe tip for use in near-field scanning optical microscopy is coated with a sheath of metal material having a plasma frequency comparable to optical frequencies. Alternatively, the sheath material has lower energy plasmons. The preferred sheath material is silver. Other preferred sheath materials are potassium, Rb, cesium, tungsten oxide, sodium tungsten oxide and Re oxide.