Abstract: A system for processing a tracking picture of an intruding object includes a first TV camera for outputting a video signal used for detecting the intruding object, a second TV camera for outputting a video signal used for monitoring a magnified tracking picture of the intruding object, and a picture processing device for processing and operating on the video signal inputted from the first TV camera and outputting a rotation control signal and a zooming control signal. The system detects a difference between the video signal inputted from the first TV camera and a reference background video signal, and a change of a detecting state from the change of the numbers of objects detected in the continuous input video signal.

Abstract: A method and apparatus for assessing the visibility of differences between two input signal sequences, e.g., image sequences is disclosed. The apparatus comprises a perceptual metric generator having an input signal processing section, a luminance processing section, a chrominance processing section and a perceptual metric generating section. The luminance processing section simultaneously processes at least two image fields, so as to provide spatio-temporal channels whose calibration is independent of pure-spatial and pure-temporal channels.

Abstract: An apparatus and method provide for interfacing a facsimile machine with a computer network. Specifically, an apparatus provides a fax interface for exchanging faxes with fax machine and a network interface for communicating with and retrieving electronic documents from a computer network. The apparatus further includes a memory for storing the electronic documents received from the computer network and the faxed pages received from the fax machine and a controller for comparing the electronic documents to the faxed pages. The apparatus may be employed to retrieve various electronic documents from the computer network in response to received faxes.

Abstract: A given point of interest in an image is defined by two properties, a local attribute, such as color, and a neighborhood function that describes a similarity pattern. The color value is not influenced by nearby background regions of the image, and functions as a descriptor for each location. The neighborhood function distinguishes locations of similar color from one another, by capturing patterns of change in the local color. The neighborhood function measures the similarity between the local color and colors at nearby points, and reduces the measured similarity values that lie beyond contrast boundaries. Through the computation of such a transform for points of interest in an image, corresponding points in other images can be readily identified.

Abstract: A method and system for printing web pages from an intra- or internet source with a client computer. The client computer is capable of accessing and viewing a web page from an intra- or internet source and has access to a program storage. The program storage device includes (i) an HTML source file containing a list of printable web pages and (ii) a program of instructions executable by the computer at a predetermined, modifiable time to print web pages from the intra- or internet source with the client computer. The program selects one or more of the printable web pages from the HTML source file, accesses the web pages selected from the HTML source file, and prints the accessed web pages. The printable web pages may be accessed on different levels of hierarchy in HTML files.

Abstract: A method and apparatus for searching, browsing and summarizing moving image data using the fidelity of a tree-structured moving image hierarchy, which can partition or divide the moving image data into predetermined units, extract a key frame or representative frame from each of the partitioned or divided units, create a tree-structured key frame hierarchy on the basis of the extracted key frames, store a fidelity value of each of the key frames, which is a sub-tree information representative degree of each key frame, at an edge between adjacent ones of the key frames and effectively and efficiently search, browse and summarize the moving image data using the stored fidelity values. Therefore, a query image or video clip desired by a user can more rapidly and accurately be searched for.

Abstract: To provide an image detection method for judging whether a specific image is present in an input image by temporarily holding the input image signal, obtaining data from the held data like a circle having a radius r, calculating the characteristic value of the circular data, judging whether the characteristic value is included in a specific pattern in accordance with the similarity between the characteristic value and a reference characteristic value, obtaining data like a circle having a radius r' different from the radius r when judging that the characteristic value is included in the specific pattern, calculating a new characteristic value, and repeating the operation of judging whether the new characteristic value is included in a specific pattern over circular data having a plurality of radiuses. Thereby, it is possible to efficiently detect the presence or absence of a specific image whose arrangement conditions such as the angle and position in an input image are unknown.

Abstract: In a method for matching two images, each consisting of a plurality of pixels and having partially overlapping contents, the degree of correspondence between the contents of the images is determined for different displacement positions representing different overlappings of the images. More specifically, a plurality of numbers are determined for each one of a plurality of said displacement positions. Each number is formed with the aid of pixel values from both images and is used to simultaneously retrieve predefined overlap assessment values for at least two of said displacement positions. The overlap assessment values retrieved are subsequently used to determine the degree of correspondence between the images for the different displacement positions. The method is carried out with the aid of a computer and can be implemented as a computer program.

Abstract: An invention for quantitatively analyzing bitmap images using computer vision is disclosed. In one embodiment, these bitmap images correspond to the rendered screens of two applications based on a same data source file. An optional preprocessing step filters non-important information for analyzing purposes such as the background, replaces each image with a blank image of the same size, and modifies non-visible attributes of a screen so their structure can be recovered during the computer vision processing. After a series of rendered screens of each application are captured, these bitmap images are interpreted using computer vision techniques to produce machine-readable visual attributes of the rendered screens. Corresponding attributes from each of the two applications are then compared to generate a set of differences, where those differences within a predefined set of deliberate design differences are ignored.

Abstract: A method and apparatus are disclosed for testing the accuracy of images generated by a computer graphics program. An output image created by a graphics program on a particular computer platform is verified by being compared to a reference image, also known as the golden image. The output image is partitioned into several checking apertures where each such checking aperture can be of either a block type or a point type. In either case, each checking aperture is made up of several pixels. One or more attributes of the output image is compared against the reference image upon execution of a test program. For pixels in a point type checking aperture, a positional tolerance is determined. Expected values of pixels in the reference image are then derived by examining the specification of the computer graphics program. The positional tolerances and the expected values, along with color tolerances, are stored in a reference file.

Abstract: A process for image analysis which includes selecting a number M of images, forming a model graph from each of the number of images, such that each model has a number N of nodes, assembling the model graphs into a gallery, and mapping the gallery of model graphs into an associated bunch graph by using average distance vectors &Dgr;ij for the model graphs as edge vectors in the associated bunch graph. A number M of jets is associated with each node of the associated bunch graph, and at least one jet is labeled with an attribute characteristic of one of the number of images. An elastic graph matching procedure is performed wherein the graph similarity function is replaced by a bunch-similarity function.

Abstract: An apparatus for determining the profile of an object provides a sensor having at least one light source which projects a sheet of light that is intersected by the object, whereby a stripe of points on the object are illuminated. The light source projects the sheet from multiple angles, so that the resulting stripe represents a continuous profile. The profile is collectively viewed by at least two optical detectors, which each have a different view than the others and provide an output signal representing its particular view of the profile. The output signals are received by an image processor, which digitizes the signals, provides filtering, and identifies (extracts) the points in each view that represent the profile. The image processor uses a predetermined set of calibration parameters to transform the coordinates of the extracted points to corresponding coordinates in a common reference system that collectively represent the profile of the object.

Abstract: The present invention includes a method, a computer product and a system each of which features fitting a spatial recognition technique to enable a microprocessor to recognize hand-written characters, referred to herein as a cipher. This is achieved by fitting the cipher to a matrix having a plurality of cells with each cell having a plurality of pixels corresponding thereto and determining whether said cipher may be equally distributed among the plurality of cells. In this fashion, a subset of the pixels of a subgroup of the plurality of cells have a sub-portion of the cipher associated therewith and it is determined whether the number of pixels associated with each portion of the cipher are equal. If not, cells have excess pixels associated with a portion of the cipher, compared to other cells in the matrix, have the excess pixels assigned to a different cell. This facilitates creating a digital word that describes the cipher that is then compared with information in a look-up table.

Abstract: An anti-counterfeit currency detection method is disclosed wherein local edge information is utilized for accurately detecting lines and curves of legitimate notes. The method can more accurately determine the location and orientation of a pattern and thus provides more reliable currency detection. A detector is trained off-line with example notes resulting in a stored template generated by recording a test pattern similar to a pattern to be tested; anchor lines are identified within said template which are further represented in subsequent test patterns. The template is rotated and shifted before matching it to the test pattern so that anchor lines align with long lines detected within the test pattern. The template and test pattern are then compared to determine whether there is a match.

Abstract: Provided is the ability to produce an orthogonal-view representation of a selected feature plane of a three-dimensional object. A plurality of images of the object are acquired, each corresponding to a distinct orientation of the object about a selected object axis. In at least one acquired image, feature points are identified in the selected feature plane as-projected into that acquired image. Feature points are associated with an orthogonal-view representation of the selected feature plane, and feature points from at least one acquired image are correlated with physical orientations on the selected feature plane based on the object orientation corresponding to that acquired image. An orthogonal-view representation of the selected object feature plane can be analyzed for a specified feature configuration even when the acquired object images are perspective-view images. This can be accomplished even through only a subset of feature points may be available in any one given image of the object.

Abstract: An apparatus for determining a similarity between model image and inputted image, which is robust even if the background of said inputted image is partly changed. The similarity is calculated from an edge direction of an inputted image, an edge direction of a model image and a weight related to the model image. The calculating steps used are (1) calculating an edge direction difference between an inputted model image and a model image, (2) calculating weights related to model image, (3) multiplying edge directions and weights, (4) calculating a sum of the multiplication, (5) calculating a sum of the weights, and (6) calculating a similarity by dividing the sum of the multiplication by the sum of the weights.

Abstract: The present invention provides a method for the calculation of a pattern similarity metric that is locally normalized with respect to illumination intensity, and is invariant with respect to rigid body preserving gray scale variations, such as scale, rotation, translation, and non-linear intensity transformations. In one aspect, the invention provides a method for comparing a model image with a run-time image so as to provide a quantitative measure of image similarity. In another general aspect of the invention, a method is provided for searching a run-time image with a model image so as to provide at least one location of the model image within the run-time image.

Abstract: A computer program product for locating first and second objects, each having substantially the same physical characteristics, and the ratio of the distance between the first and second objects and the size of each object is substantially invariant, the computer program product comprises: a computer readable storage medium having a computer program stored thereon for performing the steps of determining potential flesh regions in an image; determining valley regions in an image; performing template matching for determining a plurality of locations that give a desirable match of the object relative to the template; and performing verification for determining the likelihood of pairs of potential eye candidates at the locations determined in the performing template matching step.

Abstract: Apparatus and methods for transmitting graphical objects from a computer system to a printer. A graphical object is received from an application program. Graphical objects are stored in a computer memory coupled to the computer system. If the received graphical object matches a graphical object stored in the computer memory, an identifier identifying the matching graphical object, rather than the graphical object itself, is transmitted to the printer. Otherwise the received graphical object is stored in the computer memory and transmitted to the printer.

Abstract: An apparatus for sensing an object within an imaging field of an image pickup device includes a memory for storing data of a reference background picture of the imaging field, a device for inputting picture data from the image pickup device, a device for comparing the inputted picture data with the stored reference background picture data to detect a difference value between the inputted picture data and the stored reference background picture data, and to detect a polarity of the difference value, a converter for converting the difference value into binarization data based on a difference threshold value which is changed in response to the polarity of the difference value so that an object within the imaging field can be distinguished from an area of increased brightness within the imaging field and an area of increased darkness within the imaging field, a calculator for calculating a similarity between the inputted picture data and the stored reference background picture data, and a detector for sensing whe

Abstract: A similar change information takeout apparatus is intended to take out automatically unit containing similar change information when character row data divided in certain units are given in a multiplicity, and comprises an input unit in which character row data are entered sequentially in time series, a former data memory unit for storing character row data, a start/stop input unit for accepting an input for designating start or stop of output of similar change information from the user, a new data unit takeout process for taking out sequentially the units contained in new data, a unit comparison process for comparing the units taken out in the new data unit takeout process and the units contained in former data to see if different or same character by character contained in the units sequentially, a different character and type counting process for calculating the variety and number of different or same characters as a result of comparison in the unit comparison process, and a similarity judging process for

Abstract: Method and apparatus for aligning more than two fragments of an image to assemble the image while providing high alignment quality between each pair of overlapping image fragments. Image registration operations are performed rapidly. The disclosed method and apparatus finds application in, for example, scanning, copying, and facsimile transmission of large format documents.

Abstract: The present invention relates to a combined approach of image and template correlation, and vector correlation wherein edge detection and pattern searching are joined in a single operation to provide great flexibility without requiring extensive computational facilities. A template is devised from a sparse set of coordinate points that model only the important edges of an object to be located within an image. Different templates are dynamically reconstructed as an estimation of the object location within the image becomes more precise.

Abstract: An apparatus and method for manipulating a referenced document and a searched document containing images and descriptive text. The manipulation includes conversion of images and text into equivalent images which are standardized by utilizing predetermined design rules. The images are then compared and analyzed to determine similarities, partial similarities, and dissimilarities.

Abstract: A method for automatic matching of planar point patterns in a reference pattern with a test pattern includes course mating, calculation of a mated possibility for one point in each pattern using fuzzy relaxation, preliminary selection of mated pairs based on these possibilities, and calculation of a similarity value based upon the number of mated pairs, the mated rate of the reference and test patterns, and the average mating possibility and average distance between mated points.

Abstract: A multiresolution method and apparatus for searching of a database of images where the search is performed on compressed images, without first decompressing them. The method searches the database of compressed images first at a low resolution to obtain the relative quality of a match between a search template and a candidate image. If the match is below a particular threshold, the search is terminated without committing any further computational resources to the search. Conversely, if the match is above a particular threshold, the method enhances the resolution of the candidate image and then performs another match. As long as the relative quality of the match is above the particular threshold, the resolution of the candidate image is successively enhanced, until a match determination is made at a full resolution of the candidate image.

Abstract: In order to reduce the size of an apparatus capable of obtaining a result of correlation between images at a high speed, a multiplex selector for shifting addresses of a plurality of memories, a detector for detecting an amount stored in a memory element of one of the plurality of memories and an amount stored in a memory element obtained by shifting an address of the one of the plurality of memories, and a determination unit for obtaining correlation between outputs of the detector for each amount of shift.

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: A method and apparatus for comparing symbols extracted from binary images of text for classifying into equivalence classes. The present invention uses a Hausdorff-like method for comparing symbols for similarity. When a symbol contained in a bitmap A is compared to a symbol contained in a bitmap B, it is determined whether or not the symbol in bitmap B fits within a tolerance into a dilated representation of the symbol in bitmap A with no excessive density of errors and whether the symbol in bitmap A fits within a tolerance into a dilated representation of the symbol in bitmap B with no excessive density of errors. If both tests are passed, an error density check is performed to determine a match. The dilated representation of the bitmap accounts for various quantization errors that may occur along the boundaries of a symbol defined in the respective bitmaps. The dilation utilized preserves the topology of the symbol.

Abstract: A bitmap comparison technique that is able quickly to compare two bitmap images while discounting differences between the images likely due to noise. The bitmap comparison technique includes the operations of: comparing the first and second bitmaps, producing a difference map identifying differing bits between the first and second bitmaps, producing outline masks based on the outlines of the first and second bitmaps, identifying certain very different bits within the difference map that are to be weighted differently from the remaining bits within the difference map based upon the outline mask, and determining a comparison score to indicate the extent to which the first and second images differ by differently weighting the very different bits and the remaining bits. The certain bits are normally weighted to a lesser extent than the remaining bits when determining the comparison score so that the influence of noise in the comparison score is diminished.

Abstract: The present invention provides an image processing apparatus and method comprising a memory for storing address information of a binary registered image. An original test image is converted into a binary test image so that the ratio of the total number of black pixels to the total number of black and white pixels lies within a predetermined range. The binary test image is aligned to the binary registered image in order to compare the two images. A degree of concordance between the binary test image and the binary registered image is determined if the comparison satisfies a predetermined condition of concordance. A degree of discordance between the binary test image and the binary registered image is determined if the comparison satisfies a predetermined condition of discordance. The binary test image is determined to originate from the same object as the binary registered image according to the degree of concordance and the degree of discordance between the two images.

Abstract: A computer implemented apparatus and method for transferring information from one set or sets of memory locations to another set or sets of memory locations. The present invention has particular advantageous use within a computer system specially implemented for pattern recognition applications, such as handwriting or voice recognition. The present invention includes a system with an automatic sequencer able to sequentially generate sequential source and destination addresses and able to generate appropriate data requests to internal and external memory controllers. The present invention memory to memory transfer unit allows memory transfer operations to occur in parallel with the operation of arithmetic pipelines that process pattern recognition procedures. Therefore, using the present invention, no additional processing time is consumed by a memory transfer. Double buffering is utilized to transfer information and process information in the same time frame.

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 comparator for comparing the test image to the template. The comparator respectively 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 comparator respectively 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: A computer module is interactively trained to recognize patterns of spectral or textural features in imagery. The interactively trained module is based on techniques of knowledge-based image processing and an approach that uses interest images to provide a means of continuous feedback of module performance to the user. The user, in turn, responds by indicating where the module is making mistakes. A set of user indicated examples and counter examples form the inputs to a machine learning program called Functional Template Learning. The trained module is exportable as an independent agent to search large databases for matching patterns. Following retrieval of images having matching patterns, a user can further refine the agent's performance by indicating where mistakes have been made. The overall search tool employing the trained module is capable of prescreening and highlighting images, significantly reducing the workload of analysts attempting to detect regions or objects in imagery.

Abstract: An apparatus and method for detecting anomalies morphologically detecting regions of interest. The apparatus and method include a region of interest (ROI) detection unit which morphologically detects anomalies in two dimensional input images or morphologically detects regions of interest based on structuring elements which can be set by a user using a modified hit and miss transform which includes a thresholding feature. The thresholding feature makes it possible to vary how much the shape of features within the input images can deviate from the shape of structuring elements and still be detected as a region of interest. One application of this involves detection of abnormal cells in a cervical Pap-smear.

Abstract: A method, and related apparatus, for recognizing alphanumeric characters within a sample includes applying ideal alphanumeric character templates to the sample to identify each template which is included substantially within the sample. A template will be considered to be included substantially within the sample even if the sample has portions thereof with which the template does not substantially correspond or substantially match. That is, a template can be considered to be included substantially within the sample even if the template and the sample do not have a substantially one-to-one symmetrical correspondence. Of all of the templates which are identified as included substantially within the sample, the template having the most information (e.g., the template which requires the largest number of data points to define its shape) is selected as a recognized version of the sample.

Abstract: A method is provided for normalizing correlation data from a holographic correlator. After a sequence of object image patterns are recorded in an angularly multiplexed hologram, the entire volume hologram contents are read out with a uniform (plane wave) object beam. The plane wave readout produces an array of spots at a correlation plane, with the intensity of each spot representing the strength of the recorded hologram for the corresponding stored image pattern. The spot intensity values are accurately detected and stored as normalization factors. Holographic correlation output array values produced by parallel comparison of an input object pattern with the stored image patterns are then divided by the corresponding normalization factors. The resulting normalized correlation output accurately represents the true degree of correlation between the input object pattern and the stored image patterns by eliminating the uncertainty factor.

Abstract: A processor is provided with first and second document images. The first image represents an instance of a reference document to which instance a mark has been added. The second image is selected from among a collection of document images and represents the reference document without the mark. The processor automatically extracts from the first document image a set of pixels representing the mark. This is done by performing a reference-based mark extraction technique in which the second document image serves as a reference image and in which substantially the entirety of the first document image is compared with substantially the entirety of the second document image. Also, the processor is provided with information about a set of active elements of the reference document. The reference document has at least one such active element and each active element is associated with at least one action.

Abstract: A method, and related apparatus, for recognizing alphanumeric characters within a sample includes applying ideal alphanumeric character templates to the sample to identify each template which is included substantially within the sample. A template will be considered to be included substantially within the sample even if the sample has portions thereof with which the template does not substantially correspond or substantially match. That is, a template can be considered to be included substantially within the sample even if the template and the sample do not have a substantially one-to-one symmetrical correspondence. Of all of the templates which are identified as included substantially within the sample, the template having the most information (e.g., the template which requires the largest number of data points to define its shape) is selected as a recognized version of the sample.

Abstract: A system for inserting images into live video fields includes a method for rapidly and efficiently identifying landmarks and objects. Initially a first template, having a first pattern similar to one of the distinctive features of the object, is passed over the video field and compared to it in order to preliminarily identify at least one possible distinctive feature as a candidate. A second template is then created by taking one of the major elements of the distinctive feature candidate and extending that element all the way across the second template and then comparing it to the distinctive feature candidate. This eliminates one or more possible falsely identified features. A third template is then created having a pattern formed from another major element of said distinctive feature and extending it all the way across the third template. The third template is then likewise passed over the distinctive feature candidate and compared therewith in order to eliminate still further falsely identified features.

Abstract: The present invention provides an image pattern identification/recognition method which enables the reduction of the work load of the user required in the case of incorrect character recognition, as well as facilitating the addition of a new recognition pattern in the dictionary so that the pattern can be identified and added thereafter. A binary image pattern which has been entered is compared to each of the binary mask patterns prepared in advance for each of the recognition categories so as to identify/recognize the binary pattern and to determine the recognition category to which the binary pattern belongs.

Abstract: Images such as product images are identified as substantially similar to one or more reference images by finding pixels which have substantially the same value or values in all or substantially all of the reference images. Image erosion and/or dilation may be used in processing the reference image data to help identify pixels which can always be expected to have the same value in all acceptable images. The values associated with the corresponding pixels in a product image are combined and compared to an expected value. The values associated with pixels which correspond to regions not similar in substantially all of the reference images are combined to produce a discriminant function. The product image is identified as substantially similar to the reference image or images only if the combined values from the product image compare favorably with the expected value, and the dot product between the product image and the discriminant function is equal to a predetermined value.

Abstract: Method and apparatus for image processing for comparing fingerprints and the like. To avoid reducing the accuracy of the comparison due to the variations of the fingerprint lines when the fingerprint input is made, a method has been resolved wherein the line width of the registered fingerprint is narrowed more than the line width of the tested fingerprint. Reduction of the throughput of recording and comparison is solved by the selective utilization of the input of an activated fingerprint image, the judging of the fingerprint boundaries and the processing of only the activated division, and, additionally, in the comparison, by midway abandonment, midway decision, multi-step variation of the increment, and/or introduction of a sub-template jump search. By introducing the degree of discordance between two images, it becomes possible to compare two images based on a comparison of the black pixels of the registered image to those of the image undergoing testing.

Abstract: An apparatus for detecting errors in frames of seismic data that can be presented as images, such images being expected to exhibit an inherent symmetry about an axis, the apparatus comprising: a storage device for storing a frame of seismic data which can be presented as an image; a selection device for selecting an axis within the image as the axis of symmetry; a divider for dividing the data about the axis of symmetry to create first and second sets of data; and a comparison means for comparing the sets of data according to preselected matching criteria to determine, based on the degree of matching, whether the extent of mismatch in the frame of data is within a predetermined tolerance.