Abstract: The present invention relates to a method for enhancing the outline of shapes that occur in an original image which has been obtained by the use of an electro-optical scanner or similar device for the purposes of preparation for printing. The apparatus for reading images comprises the RAM 12 where plural sets of data having the relationships between the difference signal S-U and the enhancement signal K.multidot.(S-U) are stored. By selecting a proper set, the outline enhancement or graininess sublimation threshold setting can be set according color values and other photo aspects of the original, or according to the design of the original image and it's intended use.

Abstract: A system and method for obtaining salient contours from two-dimensional images acquired by a sensor is disclosed which process the two-dimensional images with an edge detector to produce edgels from each of the images, link the edgels into lists known as contours, compute a saliency value for each of the contours, rank the contours in decreasing order of saliency, and select certain ones of the ranked contours based on the requirements of a particular vision application.

Abstract: A color analysis and evaluation method is disclosed based upon a transformation of three primary color measurements, red, blue and green, from an x,y,z cartesian coordinate system to a rotated spherical coordinate system in which a target color vector in the cartesian coordinate system defines the primary z axis in the rotated spherical coordinate system. In one disclosed embodiment, a computerized color analysis is performed upon a finished baked product such as a pound cake with a split in the top thereof. A detector such as a color television camera measures three primary color components of each pixel of a matrix of pixels into which an image of the product is divided by the detector. The three measured primary color components are then transposed to a reference spherical coordinate system having the z axis aligned with a target color for the product. A determination is then made as to whether a measured color vector is within predetermined limits to determine if it is within an acceptable color sector.

Abstract: A method for collating independent figure elements in an object image with a reference image both including independent elements which are not interlinked with each other. The method comprises steps of defining a relative coordinate system in an image plane for the reference image and the object image on the basis of centers of gravity of two distinctive independent elements which are identified with values indicative of features of the elements; and collating pairs of independent elements between the reference image and the object image on the basis of the relative coordinate system.

Abstract: A character recognition method comprising the following steps: (1) acquiring a two dimensional array of pixels, (2) locating an unknown character in the two dimensional array, (3) computing weighted correlation coefficients between the unknown character and a trained set of characters (i.e. a font), (4) recognizing the unknown character as the trained character with the highest weighted correlation coefficient above a threshold. The weights in the correlation calculations are adjusted to place more emphasis on those areas of a trained character that distinguishes it from all other trained characters in the training set. A method for optimally adjusting these weights is described herein.

Abstract: An X-ray generator and associated method are disclosed. The X-ray generator and method are provided for inspection of objects which are conveyed into an inspection station which includes an X-ray device and image processing device. The X-ray device directs high-energy, short-pulse X-rays toward an object as the object moves into a desired position. This X-ray radiation is received by a video camera and the resulting image is then digitized and transmitted to an image processing device which analyzes the information and determines whether or not the object conforms to a predetermined standard threshold. One aspect of the apparatus includes a rejection device which will remove objects which do not pass inspection thresholds.

Abstract: A method of generating a binarized image from an original image. Error values of binarized pixels in the neighborhood of a pixel to be processed are used to modify the pixel to be processed. The error values are filtered according to weighting factors which vary as a function of the density value of the pixel to be processed independent from the density values of the pixels in the neighborhood. The weighted error values are used to modify the pixel to be processed prior to comparing the pixel to be processed to a threshold.

Abstract: In a document scanner that is operating in the binary mode, the magnitude of the signal that is generated for each picture element (PEL) of each document scan line is compared to the magnitude of a threshold signal. Either a binary 0 or a binary 1 is generated as a result of this comparison. The threshold signal magnitude is dynamically established by a method and an apparatus that provides a stored history or histogram of the various signal magnitudes that are generated for each PEL of a plurality of the document scan lines. Since the signal magnitude(s) that is provided by scanning the background, non-image, area of the document occurs the majority of the time within this histogram, the magnitude of this majority signal(s) is used to set the threshold signal magnitude. As a result, the PELS of the document's visible image are converted to a binary electronic image with a minimum loss of visual image content.

Abstract: An image restoration apparatus comprises an image display device for displaying an image estimated as a nondegraded original image, a lens system for focusing the estimated image displayed on the image display device, an image input means for photographing an image focused by the lens system and for receiving the photographed image as image information, a comparing means for comparing the input image information received by the image input means with degraded image information prepared beforehand to obtain a difference therebetween, and a restoring means for performing restoration of an original image by changing the estimated image by using a random number on the basis of the difference calculated by the comparing means.

Abstract: Locating and recognizing characters on a machine produced label, for example, is accomplished faster and more accurately by searching for first and second identifiable objects within first and second search windows, respectively. The first and second search windows are typically much smaller than the label or image being searched. The first object is the top portion of the label and the second object is the bottom portion of the label. Having located each object, a reference point is determined on each. From these two reference points, precise character or bar code locations can be predicted for reading the characters or bar codes printed on the label at predetermined locations. A character search window superimposed over each character at its estimated location is searched to read the character located thereat. If the first or second object could not be located, due to a torn label for example, an alternate object may be searched and a reference point located thereon.

Abstract: Image data processing method and system for giving area identification labels to individual areas of connected black (or white) elements included in image data are disclosed. Image data compressed by modified READ coding is successively analyzed in units of code word. Each time the image data is analyzed in units of one code word, a run table for storing a relationship between run data indicative of the position of each run and the value of a provisional label given to that run and a label table for storing a relationship between provisional labels to be later unified so as to have the same value are updated, without a wait for the expansion of the whole of the image to bit data, on the basis of a predetermined processing rule defined corresponding to the kind of a code word analyzed. When the processing has been completed for all of code words, the value of a provisional label for each run stored in the run table is corrected on the basis of the contents of the label table.

Abstract: A character recognition device includes a ruled line extracting unit supplied with image information to be recognized and operated for extracting ruled line coordinates from the image information, a character recognizing unit for recognizing characters from the region excluding frame lines extracted by the ruled line extracting unit, and a recognition result outputting unit for outputting character codes together with ruled line codes from the results obtained by the ruled line extracting unit and the character recognizing unit.

Abstract: An apparatus for scanning documents combines digitizing circuitry with a scanning platen to achieve precise selection of image areas on large documents. A vacuum system holds the document in place against a perforated scanning surface during registration and scanning. The assembly of scanning surface, vacuum chamber, and digitizing circuitry move as a unit beneath a fixed scanning apparatus. The apparatus further has a mechanical structure which allows the table assembly to rotate about a pivot axis toward an operator to reduce parallax error in mapping image areas.

Abstract: A document processing, archival storage and printout system such as for handling customer checking accounts. Original checks/documents are processed into digital image data then stored temporarily in magnetic media and transferred to optical long-term archival storage. The system retrieves and accumulates monthly groups of digital image data, then sorts one-day's worth (1/22 of 22 business-day accumulation) by account number so that printing means can print statements, each day, covering 1/22 of the total accounts existing for that month. Massive amounts of data can be accumulated and stored, for example, for 500,000 to 1,000,000 customer accounts, while the system operates to rapidly retrieve and printout sufficient customer statements each day so that each of the many customers will still receive a personal monthly updated statement during the appropriate month.

Abstract: Pattern categorization is provided by a self-organizing analog field/layer which learns many-to-many, analog spatiotemporal mappings. The field/layer employs a set of input nodes, each input node having two long term memory weights, and a set of output nodes. Each input node is for categorizing patterns with respect to a plurality of categories. The long term memory weights of an input node encode the patterns categorized by the input node. Each input node generates signals as a function of respective long term memory weights and input signals to the input node. Each input node is coupled to a different output node. Each output node receives signals generated by the respective input node and selects a category of the respective input node. The output nodes provide a mapping between plural parts of the input pattern and plural categories of the input nodes. Category selections of the output nodes are modified such that sum of the output signals from the output nodes is within a predefined range.

Abstract: To encode data representative of correlated images, a sequence of arrays of source image data is provided in which each array is representative of an image. Each array of source image data is partitioned into a plurality of source image-data blocks. For each sequence of source image-data blocks, at least one prototypical image-data block is derived from at least a portion of the source image data of the sequence. For each source image-data block, a translation coefficient and a number of scaling coefficients equal to the number of prototypical image-data blocks associated with the sequence are determined. A difference image-data block is formed for each source image-data block from elements of the source image-data block minus the associated translation coefficient minus the sum in turn of each associated scaling coefficient multiplied by elements of the corresponding prototypical image-data block.

Abstract: Disclosed is a method for improving the dynamic resolution of an imaging system. The method employs a dual sampling or exposure technique which samples light from an image intensifier over a long and a short sample interval. When the resulting measurement for the long sample interval exceeds a threshold level, the short sample interval measurement is used, multiplied by a scaling factor. Below the threshold the long sample interval is used.

Abstract: A method and an apparatus for inspecting a three-dimensional surface shape and pattern of an object under inspection by images of the surface of the object, said images being formed by reflected light rays from the object illuminated by lights. The surface of the object is illuminated by first and second lights provided on the inspected surface side of the object and at different heights from the object, images of the reflected light rays from the surface of the object illuminated by the first and second lights are acquired by an imaging unit, a first image of the surface of the object illuminated by the first light and a second image of the surface of the object illuminated by the second light, both acquired by the imaging unit, are input separately, brightness values at the same pixel points of the first and the second images are subjected to a division process to obtain a value, and by using this value as the brightness value of the pixel, a third image is formed which is used for quality evaluation.

Abstract: Apparatus systems and methods for image analysis, including identification, location and analysis of both patterns of distribution of target images and tracking of a high speed target image. An example is scanning shotgun targets. A target, representing a field of view having multiple images distributed therein, is wrapped around a transparent drum illuminated from the interior, rotating the drum in front of a high aspect ratio slit, optically folding the slit aperture into a series of adjacent overlapping zones the aggregate aspect ratio of which is low, scanning the moving target with a video camera, digitizing the camera image signals, electronically processing the signals to identify bright image areas (e.g. shot holes), processing the information to determine the image location and analyzing the location information to determine image distribution patterns.

Abstract: A luminance picture signal is applied to segmentation device (2) and a modal filter (3) to produce a region signal. Both the picture signal and the region signal are applied to a subtractor circuit (5) to produce a texture signal which is encoded in an encoder (7). The region signal is also applied to an edge mapping device (9) to produce a region list signal and an edge map of the original image. The edge map is subjected in an element prediction device (10) to a template to produce a prediction from a look-up table in a memory (12) regarding the value of an element in a fixed position adjacent the template. If the prediction is correct then a prediction error signal indicates no error while if an error exists either the prediction error signal conveys the actual value or where it may be found in a look-up table. The prediction error signal is encoded in a second encoder (14).