Abstract: A system and method for performing pattern matching to locate zero or more instances of a template image in a target image. The method first comprises sampling the template image using a Low Discrepancy sequence, also referred to as a quasi-random sequence, to determine a plurality of sample pixels in the template image which accurately characterize the template image. The Low Discrepancy sequence is designed to produce sample points which maximally avoid each other. After the template image is sampled or characterized, the method then performs pattern matching using the sample pixels and the target image to determine zero or more locations of the template image in the target image. The method may also perform a local stability analysis around at least a subset of the sample pixels to determine a lesser third number of sample pixels which have a desired degree of stability, and then perform pattern matching using the third plurality of sample pixels.

Abstract: An apparatus for recognizing characters read by a reading unit. A circumscribing rectangle of a read character is formed, and the degree of narrowness of that circumscribing rectangle is acquired. Characters having a degree of narrowness that is equal to or greater than a predetermined value are selected and blank areas are added to the circumscribing rectangle to yield a character area with a corrected degree of narrowness. The character is normalized by converting the character area to a specified size, and is recognized based on the normalized character. It is therefore possible to normalize even characters significantly elongated vertically or horizontally for easier recognition and to group their character patterns.

Abstract: A method and apparatus to detect marks of a specified shape and color on a document and to extract a specified pattern consisting of a number of these marks. An RGB color signal is sent, color by color, by way of image input unit 12 to binary processing unit 13. If the marks are yellow, the B signal for the pixels constituting the marks will be smaller than the other signals. The B signal is sent to mark shape extraction unit 13a and threshold processing is executed. Simultaneously, R, G and B are sent to mark color extraction unit 13b, and the fact that the marks are yellow is detected. AND element 13c finds the logical product of these processes, and a binary image is generated from which only the yellow pixels have been extracted. This binary image is transmitted to mark location detection unit 15 by way of storage unit 14. The shapes of the marks are matched against reference data and location data are extracted for the marks.

Abstract: A face recognition system is provided comprising an input process or circuit, such as a video camera for generating an image of a person. A face detector process or circuit determines if a face is present in a image. A face position registration process or circuit determines a position of the face in the image if the face detector process or circuit determines that the face is present. A feature extractor process or circuit is provided for extracting at least two facial features from the face. A voting process or circuit compares the extractor facial features with a database of extracted facial features to identify the face.

Abstract: A system and method for determining offset errors between line and pixel coordinates of landmarks in a digitized image generated by an imaging system disposed on a spacecraft and line and pixel coordinates predicted by a mathematical model of the imaging system using landmark geodetic coordinates on the Earth. The system and method use landmarks in symbolic form, such as perimeters of lakes and islands that are stored in a database. A digitized image generated by the satellite-based imaging system is processed to extract a patch of the image containing a landmark. The image patch is then upsampled (magnified). The landmark boundary is processed using a mathematical model of the imaging system to generate absolute coordinates of the boundary pixels, which are upsampled and rasterized to produce a landmark mask.

Abstract: A computer controlled optical system for angular alignment of structures which uses an optical imaging system input to a computer to synthesize regularly spaced reference and test line patterns replicated from binarized edges taken from the optical images of structures to be aligned. The two patterns are then combined resulting in an interference pattern in which the spacing of moire fringes indicates the amount of movement of the structures required to bring them to a predetermined alignment. All of the procedures are observable on the interference display. These moire fringe pattern are a quantitative measure of angular alignment between the reference and test images and of the edges from the structures from which they were taken. An operator programmed and controlled template automatically processes moire edge alignment measurements. An automatic template processor generates a test sequence template from the chosen areas for measurement by the operator.

Abstract: A special record format used for commanding light pattern shapes and addressable light pattern shape generator. The command format includes a first part which commands a specified gobo and second parts which command the characteristics of that gobo. The gobo is formed by making a default gobo based on the type and modifying that default gobo to fit the characteristics.

Abstract: To reliably and simply prevent the output of specified image data, a unit to detect a pattern is installed in one of the devices in a system consisting of an image scanner 1, control device 2 (a personal computer) and printer 3 which are connected consecutively. This pattern detection unit contains a unit which detects a specified pattern from the mass of image data representing a document. Once this pattern has been found, either the subsequent image data are altered before being output or output of the image data is interrupted. When the system starts up, the control unit for each device checks whether that device, or one of the other devices in the system, has a pattern detection unit as described above. If it determines that no device has such a unit, it halts operation.

Abstract: A method of determining an indicator for the presence of an image is disclosed in which the image is comprised of a two dimensional array of pixels. The method finds particular use in the prevention of fraudulent copying of images, such as bank notes using colour laser copiers and printers. The method firstly imposes a grid of cells over a portion of the image. Then a representative value for each of the cells is determined. Next, a measure of the difference between the representative values of adjacent cells of the grid is determined over the whole of the grid. The first three steps are then repeated for substantially all possible positions of the grid of cells to determine a grid position having an overall maximum positional difference. Finally, the grid position and the representative value for each of said cells is used as a representation of the image. The representation can thus comprise a template used in an image detection system.

Abstract: An apparatus and process for inspecting print material having a reference template generator circuit, a defect detector circuit for comparing a target image to a reference image and a defect rejecter circuit for determining the acceptability of the target image. The reference template generator circuit is made up of a pixel intensity range circuit, a pixel classifier circuit and a reference address circuit. Each pixel data of the target image is compared, according to the pixel classification of the pixel data, to each pixel data of the reference image. The flaw severity code is generated from the reference pixel comparison, which is then weighted and accumulated for the image area. The weighted sum of the occurrences of the flaw severity codes is compared with the user specified threshold for determining the acceptability of the test sample. Each entry of the look-up table contains a corresponding flaw severity code.

Abstract: The image correlation method and apparatus correlates or matches a test image with a template. The image correlation apparatus includes an image processor for partitioning the template into a number of labels, for determining the total number of pixels N.sub.T which form the template and for determining the number of pixels N.sub.i which form each of the labels i. The image correlation apparatus also includes comparison means for comparing the test image to the template. The comparison means determines, for each predetermined gray level j, the number of pixels of the test image N.sub.j,i representative of a predetermined gray level j which correspond to a predetermined label i of the template. The comparison means also determines, for each predetermined gray level j, the number of pixels of the test image N.sub.j representative of a predetermined gray level j which correspond to the template.

Abstract: The processing time and memory in converting scanned images to a hardware scription language is significantly reducer by using a "peephole" method to examine only partial images of features of the scanned image that is to be converted. The present method generated minimized feature templates (MFTs) by systematically removing all non-feature image pixels. In particular, the method and apparatus of the present invention establishes pattern estimates from image samples, eliminates unnecessary image pixels using off-line statistical analysis, and extracts feature templates from the larger size image patterns. Pattern recognition is then conducted by processing a few pixels of the unknown image pattern. An application of this technique is to recognize drawing symbols, such as, AND, NAND, OR, NOR, XOR, and XNOR gates, buffers, inverters, registers, and I/O pins from scanned images of electronic drawings.

Abstract: Methods, systems, and computer program products are provided for automatic image recognition of standard shapes which include a core-based deformable loci skeletal grid used to define and represent an object via a model template. The template includes deformable segments, the changes of which are measurable against the deformed model corresponding to an object in a subsequent image. Statistical correlation techniques optimize the match to further refine the shape of the subsequent image.

Abstract: When an image of an object is recognized, a template having a size smaller than the object is used. First, operation reference points C.sub.s (X.sub.m, Y.sub.n) are set at steps smaller than lengths along the X and Y directions of an image of an object to be recognized, and a template T(x.sub.i, y.sub.i) is set at the operation reference points C.sub.s (X.sub.m, Y.sub.n). Then, an equation, ##EQU1## is operated based on the template T(x.sub.i, y.sub.i) and the image data P(X, Y). Next, candidate points are determined as points having M(X.sub.m, Y.sub.n) larger than a reference value. Then, a position of the image of the object is determined according to the candidate points. Even if an image to be recognized is inclined, the position and the slope of the image can be detected at a fast speed even for image data with noise.

Abstract: A process and associated apparatus can automatically read a dot matrix code marking on an article even when the marking has been partly obscured or damaged in some way. The process comprises capturing a video image of the dot matrix code and correlating points on the captured image with a mask of expected dot appearance, using a known two-dimensional correlation technique. However, the correlation is performed in accordance with certain thresholding and mask adjustment criteria to improve read reliability and map the possible dot pattern in preparation for an attempt at reading the code. After automatically reorientating the map of possible dots to a read orientation if necessary, decoding of the pattern of possible dots is then attempted by comparing it with a datum grid representing rows and columns of valid dot matrix codes. The process can output a signal affirming that a valid code has been found.

Abstract: Pixels A, B, C, D and E of a reference image X, which are suitable for use in a pattern matching operation, are automatically determined on a pattern of the reference image. The automatic determination of the pixels is done based on a change of brightness at areas around the selected pixels. Coordinates of all of the selected pixels, such as (x.sub.1, y.sub.1), (x.sub.2, y.sub.2), (x.sub.3, y.sub.3), (x.sub.4, y.sub.4) and (x.sub.5, y.sub.5) are stored in memory. A scan of the image to be detected is done at a fixed offset area for each of the selected pixels. The values of the difference in brightness between the selected pixels and the scanned points are accumulated. The minimum value of the accumulated difference is detected, and the coordinate of the scanned point which provides the minimum accumulated difference is determined as a deviation of the detected image from the reference image.

Abstract: A method for entering and manipulating spreadsheet cell data is described. The present invention provides a method for determining the target cell for written information and for scaling the information to fit within the boundaries of the target cell. A multi-tiered character recognition scheme is used to improve the accuracy and speed of character recognition and translation of handwritten data. The original handwritten data is preserved so that either the translated data or original data may be displayed. The present invention also provides for improved editing of cell entries by allowing a plurality of editing tools to be selected. Manipulation of blocks of data can be accomplished with simple gestures. Arithmetic, statistical and logical functions can be invoked with a single command.

Abstract: A non destructive method and apparatus generates unique characterization data for a painting. The method and apparatus spatially correleates points in a photograph of the painting with data defining a physical property of the viewing surface or of the internal structure of the painting.

Abstract: An image processing device is disclosed which can reliably detect documents which may not be copied, which has a very small probability of false positives, and which can handle newly issued documents which may not be copied. In addition, an image processing scheme is disclosed which would accurately obtain the location of a pattern indicating a document to be detected; which would extract that pattern; which would determine correctly whether it is the right pattern; and which would make it difficult for anyone to tamper with the pattern. To achieve this, a specified pattern is imprinted before use on a document which may not be copied, and a previously specified number of positioning marks and dummy marks which identify the location of the pattern may also be imprinted. The device is able to detect a wide variety of documents and perform the required processing at high speed.

Abstract: A stroke-order-free and stroke-number-free method for the on-line recognition of Chinese characters comprising the steps of: (a) inputting a handwritten input script on an on-line basis, (b) preprocessing the input script to reduce a number of possible matching template characters; (c) performing basic stroke recognition using a database of basic strokes to identify all possible basic strokes contained in the handwritten input script; (d) performing stroke correspondence using a database of stroke correspondence rules to find matching strokes in template characters for the strokes contained in the handwritten input script; and (e) performing computation of discrimination functions using a database of character patterns and a database of spatial relationships between strokes of characters to find one or more template characters with minimum error.

Abstract: A SLAVE coprocessor connectable to a MASTER processor and cooperating with at least one external memory. The coprocessor presents an internal memory communicating with an internal data line extending between a first and second interface circuit communicating respectively with the processor and the external memory; the internal memory contains a character composed of a rectangular matrix of pixels corresponding to a digitized alphanumeric character; and the coprocessor presents an electronic arithmetic unit for comparing a character from the internal memory with a reference matrix from the external memory.

Abstract: A cache is used to provide additional information to a recognition system about what words are likely to be used in a current context. Words are added to the cache only if they are explicitly corrected by the user. If the user forgets to correct (or postpones correcting) a recognition error, nothing is added to the cache, and therefore, there is no negative impact on the recognition system.

Abstract: A software/firmware based method for determining gray level values to be assigned to a binary pixel image initially creates a linked list of templates, each template associated with a gray level value to be assigned to an image pixel value. Each template may include a central pixel value and plural neighbor pixel values or just plural neighbor values. However, each template is a listing of entries of template pixel values that are generally arranged in increasing concentric circles about the central pixel value. Each template entry includes a field for a template pixel value, a field for a template pixel address and a link field for a next template entry, in the event of a non-match determination between the associated template pixel value and an image pixel value. The method selects an image pixel window from the binary image and compares a central image pixel in the window with a logically, correspondingly located template pixel value.

Abstract: A method enables a dithered representation of an image to be converted to a continuous tone representation of the image. This undithering method compares the regions of dithered representation with sections of dither patterns to determine an appropriate continuous tone representation from which the dithered image could have resulted. When the continuous tone representation is found the region is converted into a continuous tone representation. This procedure may be spatially iterated on a pixel by pixel basis until all of the initial representation of the image is converted into a continuous tone representation. The procedure may also be modified to account for line art. Furthermore, the procedure may be applied to produce adjustments in the size of the picture including miniaturizations and enlargements of the original image. In accordance with this application, dithered portions of the original image are redithered in a manner to produce a alternative size output image.

Abstract: The present invention provides a handwritten character recognition method in which neighboring points are connected by straight lines and the segments in-between points are then interpolated using broken lines. Each broken line is equally divided into k segments, and a number is assigned to each division point therein according to the time sequence of the respective point. A (k+1)N.times.N (N is the number of strokes) matrix P consisting of the elements of row (k+1)(n-1)+1 and column k(m-1)+j comprising angle .OMEGA. formed by the straight line connecting division point i of stroke n and division point j of stroke m and the straight line connecting division point i of stroke n and division point j+1 of stroke m, is then calculated. The distance di between handwritten character Cx and matrix Qi (i=1-M) formed in the same manner from M templates corresponding to characters possessing the same number of strokes as character Cx, is calculated as the sum of the squares of each element of matrix P-Qi.

Abstract: Task order system 10 dispatches task orders throughout user network 10N concerning user record documents. Task order 10T is carried on an input medium suitable for scanning such as ordinary paper, and specifies both the dispatch task and the record document. The task order contains an input task identification (ID) 22 and record document index symbols 24. Scanner 10S receives the input medium and scans the task order thereon to obtain pixel images of the input ID and the index symbols. Classifier 10C receives the pixel images from the scanner for classification into an input ID code, and index codes. Template library 10L has a task ID template which defines the corresponding each task ID. Selector 11S selects the template code which matches the input ID code. Display device 10D notifies the user if the selection was unsuccessful. The user may correct the input ID code or manually select a template code from user keys 10K.

Abstract: A method of processing an image including the steps of: locating within the image the position of at least one predetermined feature; extracting from the image data representing each feature; and calculating for each feature a feature vector representing the position of the image data of the feature in an N-dimensional space, such space being defined by a plurality of reference vectors each of which is an eigenvector of a training set of like features in which the image data of each feature is modified to normalize the shape of each feature thereby to reduce its deviation from a predetermined standard shape of the feature, which step is carried out before calculating the corresponding feature vector.

Abstract: An aircraft docking system is installed at a point of convenience within the dock in full view of the approaching aircraft and in view of the pilot, and has a display which is readily seen by the pilot through the cockpit windshield as the aircraft proceeds along a taxiway and during a turning of the aircraft from the taxiway into the dock. Operation of the docking system is based on the use of feature extraction templates of three-dimensional images of candidate aircraft for the dock. An image of the approaching aircraft is obtained by use of ranging circuitry of an optical laser radar. By comparing the image with the rotated and scaled model, the image can be interpreted to show the present location and orientation of the approaching aircraft. As the image is updated by the radar, the aircraft travel path, as represented by successive locations and orientations of the aircraft, is determined and is compared with the required travel path for the specific aircraft.

Abstract: A pattern matching system includes a circuit that matches an input image with a template based on a correlation function. The circuit has a structure which makes it particularly suitable for implementation in Large Scale Integration (LSI) technologies. A pattern matching circuit according to this invention sets up a threshold value of a correlation coefficient and evaluates the following formula,E=N.sup.2 {.SIGMA.(f.sub.i -f.sub.m) (g.sub.i -g.sub.m)}.sup.2 -p.sub.th.sup.2 N.sup.2 .sigma..sub.f.sup.2 .sigma..sub.g.sup.2where N is a number of input data points. f.sub.1 is an input data point. f.sub.m is a mean value of input data points. g.sub.1 is a template data point. g.sub.m is a mean value of template data points. .sigma..sub.f is a standard deviation of input data points and .sigma..sub.g is a standard deviation of template data points in a range over a threshold value.

Abstract: In a method for producing geometric masks, a color image is displayed on a monitor. Color samples are taken from the image with a coordinate input unit and are deposited in a color sample memory. Color correction values that modify colors identified by the color samples can be input. A color computer serves the purpose of fast calculation of the color correction. For generating geometric masks, the color computer is loaded with an extreme value at the locations that are identified by the color samples. For that purpose, a table in the color computer is loaded with the extreme value at the corresponding location. The color image data are then applied to the color computer and the output data of the color computer are compared to a threshold. When the result is positive, a location in the mask memory is set. After all image pixels have been compared in this way to the thresholds, the mask memory contains a geometric mask at all locations that belong to the color samples.

Abstract: A method and apparatus (62) for processing images in symbolic space are disclosed. The method may include the steps of capturing a visual image (40), converting the visual image to a digital image (42), decomposing the digital image into image primitives (44), and processing the image primitives in symbolic space (46). The processing of the image primitives may include grouping the image primitives to form higher level image primitives, isolating portions of an image, filtering noise, identifying portions of an image, or adding to a knowledge base. The system or apparatus (62)for processing an image in symbolic space may include a camera (64), a video-to-digital converter (66), and a computer (68) with appropriate memory and software; the processing may include grouping the image primitives to form higher level image primitives, isolating portions of an image, filtering noise, identifying portions of an image, or adding to a knowledge base.

Abstract: A technique for detecting defects in stationary products or in products moving on a production line (102, FIG. 1) by analyzing their images uses a matrix or line-scan camera (104, FIG. 1) for taking images of products (102). The product's dimensions are measured with accuracy and the sizes and positions of their surface defects are determined. The technique is much faster and more accurate than current techniques and is based on an analysis of the histogram vectors of the full image (FIG. 5). A carefully selected template image (A1, FIG. 2) composed of templates (B1, FIG. 2) is created and saved in the memory of a computer (106, FIG. 1). The method also includes the steps of creating and saving a histogram vector of the template image loading Look-Up Tables with a shifting and quantizing function for the image gray levels saving a product image in memory to be superposed onto template image (FIG.

Abstract: An image processing apparatus includes an input unit for inputting images, and storage units for storing the images input from the input unit. The storage units store a plurality of images having common partial images. The apparatus further includes a determination unit for determining positions of the respective overlapped common partial images of the plurality of images stored in the storage units, a synthesis unit for forming an image by synthesizing the plurality of images according to a result of the determination by the determination unit, and an output unit for outputting the image synthesized by the synthesis unit.

Abstract: A method and apparatus detects function words in a first image of a scanned document without first converting the image to character codes. Function words include determiners, prepositions, articles, and other words that play a largely grammatical role, as opposed to words such as nouns and verbs that convey topic information. Non-content based morphological characteristics of image units are predetermined as well as the presence or omission of character ascenders and descenders in image units. Predetermined characteristics of function word image units are compared with the image units of an image and when a match occurs, the image unit is identified as a function word. Conversely when no matching characteristics occur, the image unit is identified as a non-function word. Additionally, image units are classified and identified as containing only upper case characters, only lower case characters, only digits, and mixed character types.

Abstract: A process for synthesizing a reference image of an object by applying real image characteristic element data to characteristic elements of a descriptive (CAD based generated) image. The process includes preprocessing the descriptive image data by identifying characteristic elements of the object in the descriptive dam and determining the coordinates of the characteristic elements. The first format of the descriptive image data and the coordinates of the characteristic elements are subsequently changed to an equivalent descriptive image and equivalent coordinates of the characteristic elements having a second format.