Patents Examined by Manav Seth
  • Patent number: 10885399
    Abstract: A method and apparatus for automatically performing medical image analysis tasks using deep image-to-image network (DI2IN) learning. An input medical image of a patient is received. An output image that provides a result of a target medical image analysis task on the input medical image is automatically generated using a trained deep image-to-image network (DI2IN). The trained DI2IN uses a conditional random field (CRF) energy function to estimate the output image based on the input medical image and uses a trained deep learning network to model unary and pairwise terms of the CRF energy function. The DI2IN may be trained on an image with multiple resolutions. The input image may be split into multiple parts and a separate DI2IN may be trained for each part. Furthermore, the multi-scale and multi-part schemes can be combined to train a multi-scale multi-part DI2IN.
    Type: Grant
    Filed: July 23, 2018
    Date of Patent: January 5, 2021
    Assignee: Siemens Healthcare GmbH
    Inventors: S. Kevin Zhou, Dorin Comaniciu, Bogdan Georgescu, Yefeng Zheng, David Liu, Daguang Xu
  • Patent number: 10869611
    Abstract: This invention relates to a system that adaptively compensates for subject motion in real-time in an imaging system. An object orientation marker (30), preferably a retro-grate reflector (RGR), is placed on the head or other body organ of interest of a patient (P) during a scan, such as an MRI scan. The marker (30) makes it possible to measure the six degrees of freedom (x, y, and z-translations, and pitch, yaw, and roll), or “pose”, required to track motion of the organ of interest. A detector, preferably a camera (40), observes the marker (30) and continuously extracts its pose. The pose from the camera (40) is sent to the scanner (120) via an RGR processing computer (50) and a scanner control and processing computer (100), allowing for continuous correction of scan planes and position (in real-time) for motion of the patient (P).
    Type: Grant
    Filed: December 11, 2017
    Date of Patent: December 22, 2020
    Assignees: The Queen's Medical Center, The University of Hawaii, The Medical Collene of Wisconsin, Inc., UWM Research Foundation, Inc.
    Inventors: Thomas Michael Ernst, Thomas Edmund Prieto, Brian Stewart Randall Armstrong
  • Patent number: 10867372
    Abstract: An electronic device according to embodiments may include: a communication module; a first processor configured to correct images, and a second processor operatively connected to the communication module and the first processor. The second processor is configured to control the first processor to correct a first image set obtained from a database, select at least one image from the first image set based at least on a correction result of the first image set, and transmit feature information on the selected at least one image to the first processor. The first processor is configured to obtain a second image set from the database based on the feature information, and generate a third image set by correcting the second image set. Various other embodiments may be possible.
    Type: Grant
    Filed: January 17, 2019
    Date of Patent: December 15, 2020
    Assignee: Samsung Electronics Co., Ltd.
    Inventors: Sungjun Lim, Hyunhee Park, Changsu Han, Youngjo Kim, Jaegon Kim, Hongseok Yang, Kwangtai Kim
  • Patent number: 10863660
    Abstract: A component picked up by a suction nozzle of a component mounting machine is imaged by a camera, the captured image is processed by an image recognition system to recognize the component, the image is determined to be normal or abnormal based on the recognition result, and in addition to classifying the image as a normal image or an abnormal image and storing the image in a storage device, component mounting boards, unloaded from a component mounting machine, are inspected with an inspection device. A stored image reclassification computer acquires the inspection result from the inspection device, reclassifies the normal image stored in the storage device, based on the inspection result, as an image whose determination as a normal image is suspect or as an image whose determination as a normal image is not suspect, and then stores the normal image in the storage device.
    Type: Grant
    Filed: November 14, 2016
    Date of Patent: December 8, 2020
    Assignee: FUJI CORPORATION
    Inventors: Kenji Sugiyama, Hiroshi Oike, Shuichiro Kito
  • Patent number: 10860830
    Abstract: A display panel including a display function layer displaying an image in a display area, a cover member including a first face and a second face disposed on an opposite side of the first face and faces the display panel, and a fingerprint sensor including a plurality of first detection electrodes disposed in a sensor base and configured to detect unevenness of an object brought into contact with or approaching the first face of the cover member, a shield electrode disposed to face the plurality of first detection electrodes and configured to suppress a change in capacitance between the first detection electrodes and the shield electrode, and switching elements disposed in correspondence with the first detection electrodes, the fingerprint sensor being arranged between the cover member and the display panel and arranged to overlap with the display area when viewed in a direction perpendicular to the first face are included.
    Type: Grant
    Filed: September 19, 2018
    Date of Patent: December 8, 2020
    Assignee: Japan Display Inc.
    Inventors: Toshinori Uehara, Yoshitoshi Kida, Hiroshi Mizuhashi, Hayato Kurasawa, Yuji Suzuki, Shoji Hinata, Toshiya Yamazaki
  • Patent number: 10846922
    Abstract: A method and device for displaying a two-dimensional image of a viewed object simultaneously with an image depicting a three-dimensional geometry of the viewed object using a video inspection device is disclosed. The video inspection device displays a two-dimensional image of the object surface of a viewed object, and determines the three-dimensional coordinates of a plurality of surface points. At least one rendered image of the three-dimensional geometry of the viewed object is displayed simultaneously with the two-dimensional image. As measurement cursors are placed and moved on the two-dimensional image, the rendered image of the three-dimensional geometry of the viewed object is automatically updated.
    Type: Grant
    Filed: October 11, 2018
    Date of Patent: November 24, 2020
    Assignee: General Electric Company
    Inventors: Clark Alexander Bendall, Michael Melvin Ball
  • Patent number: 10846556
    Abstract: A vehicle insurance image processing method includes: acquiring a vehicle insurance image; processing the vehicle insurance image by using a preset image classification algorithm, to determine at least one category label of the vehicle insurance image; and storing, based on classification, the vehicle insurance image in a corresponding usage scenario according to the at least one category label.
    Type: Grant
    Filed: December 18, 2019
    Date of Patent: November 24, 2020
    Assignee: Advanced New Technologies Co., Ltd.
    Inventors: Jinlong Hou, Haitao Zhang, Xin Guo, Juan Xu, Jian Wang, Yuan Cheng, Danni Cheng
  • Patent number: 10831264
    Abstract: Systems and methods for eye image set selection, eye image collection, and eye image combination are described. Embodiments of the systems and methods for eye image combination can include assigning an eye pose region of a plurality of eye pose regions to each eye image and determining a representation of an eye pose in each eye image. The set of the determined representations can be fused to generate a hybrid eye image, which can in turn be used to generate an iris code associated with a confidence score.
    Type: Grant
    Filed: January 17, 2017
    Date of Patent: November 10, 2020
    Assignee: Magic Leap, Inc.
    Inventor: Adrian Kaehler
  • Patent number: 10795256
    Abstract: The present disclosure provides an optical compensation film, a photomask, and an exposure apparatus. The optical compensation film includes a first region of the optical compensation film and a second region of the optical compensation film. The first region of the optical compensation film is positioned to correspond to an overlapping portion of the prisms, and is configured to allow light to pass therethrough and impinge on the overlapping portion of the prisms. The second region of the optical compensation film is positioned to correspond to a non-overlapping portion of the prisms, and is configured to allow light to pass therethrough and impinge on the non-overlapping portion of the prisms. Light transmittance of the first region of the optical compensation film is greater than light transmittance of the second region of the optical compensation film.
    Type: Grant
    Filed: December 26, 2018
    Date of Patent: October 6, 2020
    Assignee: Wuhan China Star Optoelectronics Technology Co., Ltd.
    Inventor: Xing Xiong
  • Patent number: 10789450
    Abstract: An optical imaging device for imaging a biometric input object is disclosed. The optical sensor includes an array of sensing elements. The optical sensor is configured to read a subset of sensing elements in the array of sensing elements; analyze the reading of the subset of sensing element to determine if one or more sensing elements of the subset is saturated; alter an operating point of the optical imaging device; and image the input object.
    Type: Grant
    Filed: October 20, 2017
    Date of Patent: September 29, 2020
    Assignee: Synaptics Incorporated
    Inventors: Robert J. Gove, Alvin Jee
  • Patent number: 10789485
    Abstract: A process for improving situational awareness at an incident scene includes first receiving audio or video of a user and detecting, in one or both of the audio and video, an instruction directed to a crowd. A compliance metric associated with the instruction is then accessed and one or more available second cameras having a field of view that incorporates a current location of the crowd are identified. One or more video streams from the second cameras are received that include the crowd. From the one or more video streams, determinations are made that the crowd as a whole meets a minimum level of compliance with the instruction and that a level of compliance of a particular other user in the crowd has fallen below a threshold level of compliance. In response, a computing device performs a noncompliance action.
    Type: Grant
    Filed: November 7, 2016
    Date of Patent: September 29, 2020
    Assignee: MOTOROLA SOLUTIONS, INC.
    Inventors: Maciej Stawiszynski, Pawel Jurzak, Lukasz Stanislaw Krecioch, Piotr Jan Kus
  • Patent number: 10783401
    Abstract: A method for generating black-box adversarial attacks on video recognition models is provided, comprising a) passing input video frames into a public image model, to obtain pixel-wise tentative perturbations; b) partitioning the tentative perturbations into tentative perturbation patches; c) estimating the rectification weight required for each patch, via querying the target video model; d) applying the patch-wise rectification weight on the patches, to obtain the rectified pixel-wise perturbations; e) applying one step projected gradient descent (PGD) perturbation on the input video, according to the rectified pixel-wise perturbations; and f) iteratively performing steps a)-e) until an attack succeeds or a query limit is reached. Systems and networks therefor are also provided.
    Type: Grant
    Filed: February 23, 2020
    Date of Patent: September 22, 2020
    Assignee: FUDAN UNIVERSITY
    Inventors: Yugang Jiang, Linxi Jiang, Xingjun Ma
  • Patent number: 10776658
    Abstract: An electronic device associates first information and at least a first portion of a first image, and uses a second image that includes a portion corresponding to at least the first portion of the first image to access the associated first information.
    Type: Grant
    Filed: January 11, 2019
    Date of Patent: September 15, 2020
    Assignee: Mobile Acuity Limited
    Inventors: Anthony Peter Ashbrook, Mark William Wright
  • Patent number: 10769410
    Abstract: Augmented reality (AR) based component replacement and maintenance may include receiving a first wireless signal from a pair of AR glasses worn by a user. An image of a component viewed by the user may be analyzed and compared to a plurality of images of components stored in a database that includes information associated with the plurality of images of the components. Based on a match of the image of the component viewed by the user to one of the plurality of images of the components stored in the database, the component viewed by the user may be identified. An inventory of the identified component may be analyzed to determine whether a supplier includes the identified component in stock, and in response to a determination that the supplier includes the identified component in stock, an estimated time of delivery of the identified component to the user may be determined.
    Type: Grant
    Filed: December 26, 2018
    Date of Patent: September 8, 2020
    Assignee: ACCENTURE GLOBAL SERVICES LIMITED
    Inventors: Ping Guo, Daniel Kaplan, Sunny M. Webb
  • Patent number: 10771674
    Abstract: An information processing system configured to perform predetermined processing based on out-of-focus information corresponding to a first area of an image, the out-of-focus information corresponding to a difference in focus between the first area of the image and a second area of the image that is different from the first area of the image.
    Type: Grant
    Filed: August 30, 2019
    Date of Patent: September 8, 2020
    Assignee: Sony Corporation
    Inventors: Toshiki Ono, Ryuji Shibata
  • Patent number: 10762342
    Abstract: According to one embodiment, an electronic apparatus performs a character recognition process, uses, if a stroke in a first area of the first handwritten document and a stroke in a second area are the same, a character recognition result of the first handwritten document, and performs, if a stroke in the first area and a stroke in the second area are different, the character recognition process for the second area including the different stroke.
    Type: Grant
    Filed: April 28, 2017
    Date of Patent: September 1, 2020
    Assignee: Toshiba Client Solutions CO., LTD.
    Inventor: Chikashi Sugiura
  • Patent number: 10762610
    Abstract: Disclosed is a method for determining filter coefficients. The method includes: obtaining the coefficients of a target filter and calculating the response of the target filter; computing according to collected data and/or predetermined data in accordance with a first data pattern so as to have the response of a first filter approximate to the response of the target filter and thereby determine the coefficients of the first filter; and computing according to the collected data and/or the predetermined data in accordance with a second data pattern so as to have the response of a second filter approximate to the response of the target filter and thereby determine the coefficients of the second filter. Accordingly, the difference between the responses of the first filter and the second filter is insignificant and results in less negative influence; and the first and the second filters can replace the target filter to reduce cost.
    Type: Grant
    Filed: March 25, 2019
    Date of Patent: September 1, 2020
    Assignee: REALTEK SEMICONDUCTOR CORPORATION
    Inventors: Tsung-Hsuan Li, Wan-Ju Tang, Yang-Ting Chou, Shih-Tse Chen
  • Patent number: 10748023
    Abstract: A region-of-interest detection apparatus for improving detection accuracy of a region of interest in a case where a region of interest overlaps a background region, or in a case where feature amounts of a region of interest and a background region are similar to each other calculates feature amounts of regions where partial regions and a background region set in an input image overlap each other, and based on the calculated feature amounts and a feature amount of each position in the input image, calculates a foreground level of the position in the input image. Then, the region-of-interest detection apparatus detects a region of interest from the input image based on the calculated foreground level and a saliency of the position.
    Type: Grant
    Filed: November 26, 2018
    Date of Patent: August 18, 2020
    Assignee: CANON KABUSHIKI KAISHA
    Inventors: Masao Yamanaka, Takahisa Yamamoto
  • Patent number: 10740588
    Abstract: A fingerprint sensor and a method of manufacturing the same are provided. The fingerprint sensor includes an active region including driving electrodes and sensing electrodes crossing the driving electrodes, tracer regions connected respectively to the driving electrodes and the sensing electrodes, and shield layers disposed respectively on the tracer regions.
    Type: Grant
    Filed: May 4, 2017
    Date of Patent: August 11, 2020
    Assignee: SAMSUNG ELECTRONICS CO., LTD.
    Inventors: Jinmyoung Kim, Seogwoo Hong, Seokwhan Chung, Byungkyu Lee
  • Patent number: 10740901
    Abstract: A segmentation model is trained with an image reconstruction model that shares an encoding. During application of the segmentation model, the segmentation model may use the encoding and network layers trained for the segmentation without the image reconstruction model. The image reconstruction model may include a probabilistic representation of the image that represents the image based on a probability distribution. When training the model, the encoding layers of the model use a loss function including an error term from the segmentation model and from the autoencoder model. The image reconstruction model thus regularizes the encoding layers and improves modeling results and prevents overfitting, particularly for small training sizes.
    Type: Grant
    Filed: December 17, 2018
    Date of Patent: August 11, 2020
    Assignee: NVIDIA CORPORATION
    Inventor: Andriy Myronenko