Patents Examined by Aklilu K Woldemariam
-
Patent number: 7760947Abstract: The present invention provides a method of modifying a classification scheme for classifying hand-written characters to thereby make the classification scheme user dependent. The classification scheme includes a number of allographs, each allograph representing a respective style of a respective letter. The method includes obtaining at least one hand-written character, and then selecting an allograph representing the character. One or more of the allographs are then modified in accordance with the selection, to thereby reflect the fact that the allograph corresponds to a style of letter used by the user.Type: GrantFiled: November 26, 2008Date of Patent: July 20, 2010Assignee: Silverbrook Research Pty LtdInventor: Jonathon Leigh Napper
-
Patent number: 7760944Abstract: Text data is extracted from image data read by an image reader. Log data containing the extracted text data is generated. The generated log data is stored in a log management server.Type: GrantFiled: November 9, 2004Date of Patent: July 20, 2010Assignee: Canon Kabushiki KaishaInventor: Hiroshi Uchikawa
-
Patent number: 7756341Abstract: Generic visual categorization methods complement a general vocabulary with adapted vocabularies that are class specific. Images to be categorized are characterized within different categories through a histogram indicating whether the image is better described by the general vocabulary or the class-specific adapted vocabulary.Type: GrantFiled: June 30, 2005Date of Patent: July 13, 2010Assignee: Xerox CorporationInventor: Florent Perronnin
-
Patent number: 7751607Abstract: The present invention is a multi-stage detection algorithm using a successive nodule candidate refinement approach. The detection algorithm involves four major steps. First, the lung region is segmented from a whole lung CT scan. This is followed by a hypothesis generation stage in which nodule candidate locations are identified from the lung region. In the third stage, nodule candidate sub-images pass through a streaking artifact removal process. The nodule candidates are then successively refined using a sequence of filters of increasing complexity. A first filter uses attachment area information to remove vessels and large vessel bifurcation points from the nodule candidate list. A second filter removes small bifurcation points. The invention also improves the consistency of nodule segmentations. This invention uses rigid-body registration, histogram-matching, and a rule-based adjustment system to remove missegmented voxels between two segmentations of the same nodule at different times.Type: GrantFiled: November 25, 2008Date of Patent: July 6, 2010Assignee: Cornell Research Foundation, Inc.Inventors: Anthony P. Reeves, David Yankelevitz, Claudia Henschke, Antoni Chan
-
Patent number: 7751622Abstract: A method for detecting undesirable images within an image sequence, comprising the steps of receiving the image sequence; computing one or more associated image characteristics for a group of images, wherein the group of images includes at least one image from the image sequence; identifying any undesirable group of images based on the computed image characteristics; collecting undesirable groups of images into at least one collection; receiving a threshold value of a minimum number of undesirable images; and flagging as undesirable only images that belong to one or more collections that have at least the threshold value of undesirable images.Type: GrantFiled: August 22, 2005Date of Patent: July 6, 2010Assignee: Carestream Health, Inc.Inventors: Nathan D. Cahill, Shoupu Chen
-
Patent number: 7751624Abstract: A method for processing a customer request comprises receiving a request, sending the requests to a data provider; receiving a document image from the data provider, performing pattern data extraction on the document image, and preparing a customized report. Additionally, a system for processing customer requests comprises a receiving module for receiving customer requests, a data acquisition module for receiving data from plural sources, an image distribution module for distributing data to appropriate entities, a pattern data extraction module for performing pattern detection and extraction, and a data output module for formatting data.Type: GrantFiled: August 19, 2005Date of Patent: July 6, 2010Assignee: Nextace CorporationInventor: Donald J. Cantral
-
Patent number: 7734118Abstract: The invention features a method, system, and computer program product. A source bitmap image, a destination bitmap image, a source image feature outline surrounding a source region of the source bitmap image, and a destination image feature outline surrounding a destination region of the destination bitmap image are received. A point-to-point correspondence between the source image feature outline and the destination image feature outline is generated, such that the point-to-point correspondence defines a reshaping of the source image feature outline. An interpolation surface is calculated interpolating an area of the source bitmap image that includes the source region. The interpolation surface is reshaped using the point-to-point correspondence.Type: GrantFiled: November 17, 2004Date of Patent: June 8, 2010Assignee: Adobe Systems IncorporatedInventors: Krish Chaudhury, Dejan Markovic
-
Patent number: 7729560Abstract: Exemplary methods for estimating an orientation angle and a translation values for scanned images are disclosed. The methods described herein may be embodied as logic instructions on a computer-readable medium. In one exemplary implementation meaningful image information is removed from a scanned image, resulting in a transformed image that comprises alternating, parallel lines disposed at an orientation angle relative to an orthogonal axis. The orientation angle may be determined using geometric techniques or statistical correlation techniques, and statistical correlation techniques may be implemented to determine translation values.Type: GrantFiled: November 17, 2008Date of Patent: June 1, 2010Assignee: Microsoft CorporationInventors: Mustafa Kesal, Mehmet Kivanc Mihcak, Gary K. Starkweather
-
Patent number: 7729511Abstract: In order to safeguard a hazardous area, in particular the hazardous area of an automatically operating machine, images of the hazardous area are recorded with at least a first and a second image recording unit that are arranged offset from one another. The images of the scene are subjected to three-dimensional scene analyses using different three-dimensional scene analysis methods, and the hazardous area is safeguarded when at least one of the scene analysis methods detects a foreign object.Type: GrantFiled: March 23, 2005Date of Patent: June 1, 2010Assignees: Pilz GmbH & Co. KG, Daimler AGInventors: Christian Wöhler, Werner Progscha, Lars Krüger, Dietmar Döttling, Martin Wendler
-
Patent number: 7720303Abstract: Polynomial approximation based image filter methods, systems, and machine-readable media are described. In one aspect, multiple n-order polynomial approximations to a value of a target pixel of an input image are computed from multiple respective sets of m values of pixels in a neighborhood of the target pixel, wherein n?0 and m?n+1. A respective approximation error is computed from each of the computed n-order polynomial approximations. A value of a pixel of an output image is computed from the target pixel value and the computed approximation errors.Type: GrantFiled: April 28, 2004Date of Patent: May 18, 2010Assignee: Hewlett-Packard Development Company, L.P.Inventor: Renato Keshet
-
Patent number: 7720312Abstract: An image processor includes a memory that stores image layers, ?-layers associated with the image layers, and pointers linking image areas of the image layers with the ?-layer. A method for blending multiple image layers includes obtaining a transparency value for an image area by reading a pointer associated with the image area and blending the image layers using the obtained transparency value.Type: GrantFiled: February 17, 2005Date of Patent: May 18, 2010Assignee: Harman Becker Automotive Systems GmbHInventors: Michael Maier, Bernhard Broghammer, Günther Huber
-
Patent number: 7715614Abstract: A method of producing component data for use with an electronic component mounting apparatus includes imaging a component whose component data is to be produced and displaying a component image on a display, the component image on the display having an outline having at least one side edge with two corners, inputting the corners as reference points so that a first reference line passing the reference points is produced, producing second and third reference lines passing the reference points respectively so as to be perpendicular to the first reference line, inputting a point on another side edge of the outline of the component image and producing a fourth reference line passing the point on the side edge so as to be parallel to the first reference line, and producing component data by measuring a size of the component image outline on the basis of the first to fourth reference lines.Type: GrantFiled: August 19, 2005Date of Patent: May 11, 2010Assignee: Fuji Machine Mfg. Co., Ltd.Inventors: Kazumi Hoshikawa, Toru Takahama
-
Patent number: 7715643Abstract: An image processing device for processing images of background images and moving objects. A region specifying unit specifies a mixed region made up of a mixture of a foreground object component and a background object component, and a non-mixed region made up of one or the other of a foreground object component and a background object component, and outputs region information corresponding to the specifying results. A foreground/background separation unit separates the input image into foreground component images and background component images, corresponding to the region information. A separated image processing unit processes the foreground component images and background component images individually, corresponding to the results of separation.Type: GrantFiled: June 13, 2002Date of Patent: May 11, 2010Assignee: Sony CorporationInventors: Tetsujiro Kondo, Junichi Ishibashi, Takashi Sawao, Naoki Fujiwara, Takahiro Nagano, Seiji Wada, Toru Miyake
-
Patent number: 7715638Abstract: The invention relates to a method as well as a system, a device, an encoder and a decoder, and a computer software product for image processing by the method. In the invention, the number of bits is limited in the bit string of a pixel to be processed, wherein the pixel is encoded with the limited number of bits. A prediction value corresponding to said pixel is searched for. If it is found, the difference between the pixel and the prediction value is determined, to select the method for encoding the bit string of said pixel. Also, a code word is encoded in the bit string, to indicate the selected encoding method. If the prediction value is missing, the number of bits in said pixel is limited by quantizing. By means of the invention, a fixed number of bits is obtained for all encoded pixels in an image.Type: GrantFiled: January 12, 2004Date of Patent: May 11, 2010Assignee: Nokia CorporationInventor: Ossi Kalevo
-
Patent number: 7706601Abstract: An object pose estimating and matching system is disclosed for estimating and matching the pose of an object highly accurately by establishing suitable weighting coefficients, against images of an object that has been captured under different conditions of pose, illumination. Pose candidate determining unit determines pose candidates for an object. Comparative image generating unit generates comparative images close to an input image depending on the pose candidates, based on the reference three-dimensional object models. Weighting coefficient converting unit determines a coordinate correspondence between the standard three-dimensional weighting coefficients and the reference three-dimensional object models, using the standard three-dimensional basic points and the reference three-dimensional basic points, and converts the standard three-dimensional weighting coefficients into two-dimensional weighting coefficients depending on the pose candidates.Type: GrantFiled: August 27, 2004Date of Patent: April 27, 2010Assignee: NEC CorporationInventor: Masahiko Hamanaka
-
Patent number: 7702176Abstract: A method and device of feature enhancement for digital image processing. A predefined feature in one dimension of the image is searched to obtain several peaks, and each peak is either a local maximum or local minimum. Each peak is enhanced by adding an offset to increase the local maximum or decrease the local minimum. The offset of each peak is computed by calculating the difference between each peak and the preceding peak, then multiplying by a preset ratio.Type: GrantFiled: May 1, 2008Date of Patent: April 20, 2010Assignee: PixArt Imaging Inc.Inventors: Chun-Huang Lin, Jeng-Feng Lan
-
Patent number: 7702172Abstract: A method and apparatus for locating the boundary of an object. An electronic image of the object is formed, having a plurality of image pixels. Groups of the image pixels are identified that represent edge segments of the object. Patches are formed around the image pixel groups, where each patch is dimensioned and positioned to entirely contain one of the image pixel groups. A patch merge process is preformed that merges any two of the patches together that overlap each other by a predetermined amount, to form a merged patch that is dimensioned and positioned to entirely contain the two merged patches. The merge process continues for any overlapping patches and merged patches until none of the patches and the merged patches overlap each other by the predetermined amount. All the edge segments contained within one of the merged patches are associated as representing the boundary of the object.Type: GrantFiled: November 17, 2003Date of Patent: April 20, 2010Assignee: IRIS Internation, Inc.Inventor: Eric Chapoulaud
-
Patent number: 7697765Abstract: In learning for pattern recognition, an aggregation of different types of object image data is inputted, and local features having given geometric structures are detected from each object image data inputted. The detected local features are put through clustering, plural representative local features are selected based on results of the clustering, and a learning data set containing the selected representative local features as supervisor data is used to recognize or detect an object that corresponds to the object image data. The learning thus makes it possible to appropriately extract, from an aggregation of images, local features useful for detection and recognition of subjects of different categories.Type: GrantFiled: January 28, 2005Date of Patent: April 13, 2010Assignee: Canon Kabushiki KaishaInventors: Masakazu Matsugu, Pierre Cardon
-
Patent number: 7697746Abstract: The inspection system arbitrarily selects from among a plurality of optical conditions to change a distribution of reflected or diffracted light component from an object being inspected. The inspection system has a one- or two-dimensional optoelectric conversion image sensor, optically acquires an image of the object by scanning a stage on which the object is mounted or scanning the image sensor, and processes the image to check for defects in the object. Under each optical condition (illumination optical system, detection optical system, scan direction, etc.) the object being inspected is imaged and, based on the brightness distribution and contrast in the detection field of the image sensor, image sensor output correction data is generated to correct the output of the image sensor.Type: GrantFiled: October 18, 2005Date of Patent: April 13, 2010Assignee: Hitachi High-Technologies CorporationInventor: Hiroshi Kawaguchi
-
Patent number: 7689006Abstract: Method and apparatus for securing biometric data using multiple biometrics. The method includes the steps of: a) converting a first biometric to an irreversibly altered biometric using a parameterized transform function and b) prior to said conversion parameterizing a non-invertible transform function using at least one additional biometric. The apparatus is an apparatus for converting a first biometric to an irreversibly altered biometric using a parameterized non-invertible transform function comprising a computer containing a program for calculating the irreversibly altered biometric based upon input of a first biometric into the parameterized non-invertible transform function.Type: GrantFiled: August 19, 2005Date of Patent: March 30, 2010Assignee: The Research Foundation of State University of NYInventors: Venu Govindaraju, Viraj Chavan, Sharat Chikkerur