Patents Examined by D. R. Anderson
-
Patent number: 5848179Abstract: The invention provides a method of reproducing medical images so that optimal image quality for diagnosis is obtained through the use of a specific relation between digital signal values and corresponding density values in the hard copy and by means of a preliminary evaluation of the histogram of the image before reproduction.Type: GrantFiled: September 16, 1992Date of Patent: December 8, 1998Assignee: AGFA-GevaertInventor: Johan Maria Braet
-
Patent number: 5544257Abstract: A computer-based system and method for recognizing handwriting. The present invention includes a preprocessor, a front end, and a modeling component. The present invention operates as follows. First, the present invention identifies the lexemes for all characters of interest. Second, the present invention performs a training phase in order to generate a hidden Markov model for each of the lexemes. Third, the present invention performs a decoding phase to recognize handwritten text. Hidden Markov models for lexemes are produced during the training phase. The present invention performs the decoding phase as follows. The present invention receives test characters to be decoded (that is, to be recognized). The present invention generates sequences of feature vectors for the test characters by mapping in chirographic space. For each of the test characters, the present invention computes probabilities that the test character can be generated by the hidden Markov models.Type: GrantFiled: January 8, 1992Date of Patent: August 6, 1996Assignee: International Business Machines CorporationInventors: Eveline J. Bellegarda, Jerome R. Bellegarda, David Nahamoo, Krishna S. Nathan
-
Patent number: 5513271Abstract: Input image data define an input image that shows a proportioned parts graph, such as a pie chart or a whole or segmented bar graph. The input image data are used to obtain segmented feature data indicating a feature in the input image that satisfies a constraint on segments. The segmented feature data are used to obtain proportion data indicating each segment's proportion. Various criteria could be applied to find parts of the input image that form the feature. For example, for some pie charts, the segments constraint could include a circularity criterion, a center criterion, and a direction criterion, and proportions could be obtained from directions. For other pie charts, the segments constraint could include a distinct regions criterion and a circularity criterion and proportions could be obtained from directions of region sides.Type: GrantFiled: November 24, 1993Date of Patent: April 30, 1996Assignee: Xerox CorporationInventors: Satyajit Rao, James V. Mahoney
-
Patent number: 5504821Abstract: A color converting apparatus is provided with a table memory for storing pieces of color converting data defined in an output color space at color converting addresses, and a signal receiving section for receiving three signals designating an input color in a lightness-color difference space. The color converting addresses are placed at lattice points of cubic blocks formed by partitioning the lightness-color difference space in a lightness direction and first and second color difference directions. Each of the cubic blocks has the lattice points Pa, Pb, Pc and Pd at its lower plane and the lattice points Pe, Pf, Pg and Ph at its upper plane. Each of the cubic blocks is partitioned into a first triangular prism having the lattice points Pa, Pb, Pc, Pe, Pf and Pg and a second triangular prism having the lattice points Pa, Pd, Pc, Pe, Ph and Pg.Type: GrantFiled: March 31, 1994Date of Patent: April 2, 1996Assignee: Matsushita Electric Industrial Co., Ltd.Inventors: Katsuhiro Kanamori, Osamu Yamada, Hideto Motomura, Rika Hashimoto, Teruo Fumoto, Hiroaki Kotera
-
Patent number: 5490221Abstract: A process for X-ray registration and differencing results in more efficient compression. Differencing of registered modeled subject image with a modeled reference image forms a differenced image for compression with conventional compression algorithms. Obtention of a modeled reference image includes modeling a relatively unrelated standard reference image upon a three-dimensional model, which three-dimensional model is also used to model the subject image for obtaining the modeled subject image. The registration process of the modeled subject image and modeled reference image translationally correlates such modeled images for resulting correlation thereof in spatial and spectral dimensions. Prior to compression, a portion of the image falling outside a designated area of interest may be eliminated, for subsequent replenishment with a standard reference image.Type: GrantFiled: February 28, 1992Date of Patent: February 6, 1996Assignee: The United States of America as represented by the Administrator of the National Aeronautics and Space AdministrationInventors: Gary A. Ransford, Vivien J. Cambridge
-
Patent number: 5479537Abstract: In an image processing method of effecting the pattern matching of an input image and a pre-memorized reference image, the reference image is pre-memorized in memory means and the input image is input to input means. The reference image and the input image obtained from the memory means and the input means, respectively, have the same differential filters applied thereto by differentiation means, whereby a differential reference image and a differential input image are obtained. The differential reference image and the differential input image are input to differential correlation value calculation means or inner product sum calculation means, whereby the cross-correlation value or the inner product sum of these is calculated, and the local maximum value of the cross-correlation value or the inner product sum is found by image recognizing means to thereby find the relative positional relation between the reference image and the input image.Type: GrantFiled: May 8, 1992Date of Patent: December 26, 1995Assignee: Nikon CorporationInventors: Muneki Hamashima, Shinichi Okita
-
Patent number: 5467410Abstract: The present invention provides a robust technique for quickly determining whether a binary input image originated as a blank page. The technique provides reliable sensing in the presence of various image and scanner noise in the input image. In broad terms, the invention contemplates reducing the input image with a low threshold, labeling (by size) connected components (8-connected or 4-connected), and performing a threshold analysis. The threshold analysis typically entails size and numerical thresholds, taking into account the characteristic dimensions of expected types of noise. In specific embodiments, the reduction is performed as a textured reduction wherein the image is divided into tiles, and a single row of pixels in each tile is checked to see whether there are any ON pixels. If there are, the corresponding pixel in the reduced image is ON, otherwise it is OFF. Optional morphological operations are performed to remove expected sources of noise (e.g., pepper noise and thin horizontal lines).Type: GrantFiled: March 20, 1992Date of Patent: November 14, 1995Assignee: Xerox CorporationInventor: Dan S. Bloomberg
-
Patent number: 5384860Abstract: A connectivity algorithm hardware implementation system wherein a subject pixel in a raster-scanned digitized image is declared a pixel of interest, in real time, by it being part of a potential target where every pixel on a closed pixel path enclosing the subject pixel differs in intensity from the intensity of the subject pixel by a predetermined intensity level, the system including a sequential delay line array producing a plurality of sequentially delayed digitized video signals representative of the spatial relationship of the pixels in the digitized image, these signals being coupled to a magnitude comparator array along with a delayed subject pixel contrast signal produced by a threshold circuit coupled to the subject pixel intensity causing the comparator array to produce a plurality of resultant signals representative of a comparison of the video signals with the subject pixel contrast signal, which resultant signals are coupled to a path checking circuit that simultaneously identifies all predetermType: GrantFiled: August 20, 1992Date of Patent: January 24, 1995Assignee: Loral Aerospace Corp.Inventors: Robert C. Gardemal, Jr., Kurt J. Otto, Gregory A. Roberts
-
Patent number: 5381490Abstract: There is disclosed an image processing apparatus for emphasizing edge components of an image signal to be provided from an image input system.Type: GrantFiled: August 19, 1992Date of Patent: January 10, 1995Assignee: Samsung Electronics Co. Ltd.Inventor: Jae-Sub Shin
-
Patent number: 5353357Abstract: Substantially circular objects such as the ends of the filters of filter tipped cigarettes are inspected to make sure they have an acceptable appearance. The object surface to be inspected is preferably cantilevered out from the object supporting structure, and any surface of the supporting structure which faces in the same direction as the object surface is preferably made visually contrasting with the object surface. The object surface is preferably illuminated obliquely and at an angle to the axis along which an image of the object is formed. The image is subjected to "blob" analysis to determine whether it has an acceptable appearance. Objects which do not have an acceptable appearance are rejected. A sophisticated user interface provides substantial information to an operator of the system.Type: GrantFiled: December 4, 1992Date of Patent: October 4, 1994Assignee: Philip Morris IncorporatedInventors: H. Cary Longest, Jr., Barry S. Smith
-
Patent number: 5347592Abstract: A pattern judging method which is to be implemented on a computer automatically determines whether or not a pattern should be prohibited from being formed in a mask. The pattern judging method includes the steps of (a) dividing an area of the mask where a desired pattern is to be formed into a plurality of regions, (b) calculating a predetermined physical quantity for each of the regions for a case where one or a plurality of openings corresponding to the desired pattern are formed in the mask, and (c) prohibiting the desired pattern from being formed in the mask if the predetermined quantity calculated in the step (b) exceeds a threshold value for at least one of the regions.Type: GrantFiled: July 23, 1992Date of Patent: September 13, 1994Assignee: Fujitsu LimitedInventors: Hiroshi Yasuda, Satoru Yamazaki, Kiichi Sakamoto
-
Patent number: 5347598Abstract: In an image processing apparatus, an original image is extracted in the form of line segments divided by characteristic points, the extracted line segments are classified in accordance with the lengths of line segments connected to the characteristic points, and a frame is made in accordance with classification results.Type: GrantFiled: April 19, 1993Date of Patent: September 13, 1994Assignee: Canon Kabushiki KaishaInventor: Masami Ogawa
-
Patent number: 5337371Abstract: A pattern classification system includes a plurality of classification sections. Each of the classification sections includes a device for storing information of N coefficients W representing a reference pattern, a device for calculating an evaluation value V on the basis of N input signals S and the N coefficients W, the N input signals representing an input pattern, the evaluation value V representing a relation between the input pattern and the reference pattern, a device for storing information of a fixed threshold value R, and a device for comparing the evaluation value V and the threshold value R and for outputting an estimation signal depending on a result of the comparing, the estimation signal including a category signal P which represents a category. A selection section is operative for selecting one of categories in response to the category signals outputted from the classification sections and for outputting a signal Px representing the selected one of the categories.Type: GrantFiled: August 27, 1992Date of Patent: August 9, 1994Assignee: Matsushita Electric Industrial Co., Ltd.Inventors: Masaaki Sato, Motohiko Naka, Takehiko Shida, Kunio Yoshida, Mie Saitoh, Ikuo Akamine, Makoto Shimizu, Katsuhiko Fujiwara, Akira Yokouchi
-
Patent number: 5335295Abstract: System and method for scaling the size of a digital image from a first digital image to a second digital image by a specified scaling factor of G. The apparatus comprises an interpolator module configured to magnify the first digital image by a first scaling factor G1. The interpolator is further configured to output an intermediate digital image indicative of the magnified first image. The apparatus further comprises a shrinking module configured to shrink the intermediate digital image outputted by the interpolator module by a second scaling factor G2 to thereby generate the second digital image. Scaling factor G2 is a power of two.Type: GrantFiled: May 8, 1991Date of Patent: August 2, 1994Assignee: International Business Machines CorporationInventors: Bruno Ferracini, Jorge Gonzalez-Lopez, Giancarlo Pettazzi
-
Patent number: 5329595Abstract: An image processing system and method for performing optical signal processing including target recognition and filtering. The system is particularly useful in remote or mobile imaging processing systems where the data gathering capabilities exceed the ability of the system to transmit this data over conventional data links. The system includes a spectrometer which receives portions of an image and produces a plurality of spectra of individual pixels in the image. A light-sensitive detector receives the spectra and converts the spectra into electrical signals. A memory unit stores a predetermined matrix of coefficients, each coefficient corresponding to one wavelength of said spectral signal. A multiply and accumulate unit received one spectral signal and a corresponding matrix coefficient and adds the products to generate a spectral measure for each pixel in the image.Type: GrantFiled: June 5, 1992Date of Patent: July 12, 1994Assignee: TRW Inc.Inventor: Donald W. Davies
-
Patent number: 5321767Abstract: A process for examining the appearance of an examination object has the steps of taking an image of the object to obtain image data and processing the image data by an image processing apparatus. A mask is formed to determine, in the image data to be processed, a non-examination region of various configurations and, when the position of the examination object is deviated from the correct position, the position of the mask also is shifted correspondingly to enable appearance examination in varying conditions. To this end, a brightness threshold value is set in the input multi-value image data so as to define the non-examination region to be excluded from said examination region, thereby forming said mask, and the pixels having gradation levels not higher or lower than the threshold value are excluded from the examination region.Type: GrantFiled: August 28, 1992Date of Patent: June 14, 1994Assignee: Kabushiki Kaisha Komatsu SeisakushoInventor: Takahiro Murase
-
Patent number: 5311602Abstract: An image compressing circuit is comprised of: a number generating circuit which generates a respective number for each X-bit pixel in an input image such that the respective number changes in a cyclic fashion from pixel to pixel across each row of pixels in the input image; and, a translator circuit which performs a translation on each individual X-bit pixel together with its respective number to produce a corresponding Y-bit pixel, where Y is less than X. This translation is one which, when performed sequentially on several consecutive X-bit pixels of a single magnitude and their respective numbers for one cycle of the numbers, produces a series of Y-bit pixels of one or two magnitudes with an average value that increases monotonically as the single magnitude increases. By these Y-bit pixel series, the fidelity of the input image is retained even though Y is less then X.Type: GrantFiled: February 25, 1992Date of Patent: May 10, 1994Assignee: Unisys CorporationInventors: Donald J. Nieglos, Daniel A. Neuss
-
Patent number: 5309526Abstract: An image-processing system for perceptual coding of images by analyzing one or more images to determine levels of imperceptible noise and encoding a first image to produce encoded values without introducing noise exceeding a determined level of imperceptible noise is disclosed. Analyzing of images is carried out under a set of user determined conditions comprising viewing distance and lighting. The encoded image is communicated and decoded to produce representation of the first image for display. As part of displaying the representation of the first image, a halftoning process is carried out. The halftoning process may comprise the interpolation of data values of the representation of the first image as well as the addition of micro-dither to the representation. Illustrative halftoning techniques include classical halftoning, least-squares halftoning, and error diffusion halftoning. Printer model-based halftoning techniques for both least-squares halftoning and error diffusion halftoning are presented.Type: GrantFiled: October 31, 1991Date of Patent: May 3, 1994Assignee: AT&T Bell LaboratoriesInventors: Thrasyvoulos N. Pappas, Robert J. Safranek, Nambirajan Seshadri, James D. Johnston, David L. Neuhoff
-
Patent number: 5297220Abstract: An image processing system for image compression and decompression includes a template making part for producing a template from an image with respect to one target pixel in the image, the template containing a prescribed number of reference pixels selected from among pixels preceding the target pixel on a plurality of scanning lines including a current scanning line, the reference pixels having respective colors which have been detected, a storage part for storing the colors of the reference pixels in the template, and a state index detector for detecting a state index of the template from the colors of the reference pixels by referring to a state index table in which correspondence between values of the state index and states of colors of the reference pixels is pre-defined, so that the image is encoded into a compressed image and the compressed image is decoded into a reconstructed image in accordance with the thus detected state index of the template.Type: GrantFiled: August 24, 1992Date of Patent: March 22, 1994Assignee: Ricoh Company, Ltd.Inventor: Yasuyuki Nomizu
-
Patent number: 5272762Abstract: The image of a wiring pattern provided on a printed board is obtained and binarized. A square pixel operator comprised of a matrix of pixels is applied to the pattern image. Pixel trains extending in different directions are extracted from the result of applying the square operator. The numbers of pixels having the binary value "1" on the pixel trains are counted to obtain the respective sizes of the pattern image on the pixel trains. Signals representing the respective sizes are subjected to timing adjustment operation and to obtain the width of the wiring pattern on an area larger than the square operator.Type: GrantFiled: February 6, 1992Date of Patent: December 21, 1993Assignee: Dainippon Screen Mfg. Co., Ltd.Inventors: Yoshinori Sezaki, Tetsuo Hoki, Takao Kanai, Hitoshi Atsuta