Patents Examined by Randolph I Chu
  • Patent number: 11348265
    Abstract: A method and system for creating a point cloud are disclosed. A first image is captured by a first camera sensor and a second image is captured by a second camera sensor. The first and the second image have an area of overlap. Location of the first camera relative to the second camera is predetermined and each of the first and the second camera is having a field of view of more than 180 degrees. Based on the area of overlap, the first and the second image are stitched to create a composite stitched image. In one aspect, depth information from the area of overlap is extracted based on the predetermined location and a point cloud is created from otherwise to be discarded image data in the area of overlap.
    Type: Grant
    Filed: September 14, 2018
    Date of Patent: May 31, 2022
    Assignee: Snap Inc.
    Inventors: Simon Saito Haagen Nielsen, John Christopher Collins, Allan Joseph Evans, Graham Shaw, Vikas Gupta
  • Patent number: 11328390
    Abstract: A visualization system including a video processing apparatus (VPA) including a processor; and memory having processing instructions stored therein and executable by the processor, the processing instructions operable to, when executed by the processor: determine an amplification gain level applied by the image sensor; determine, based on the amplification gain level, a denoising level and a corresponding sharpening level; and process image data to denoise and sharpen an image corresponding to the image signals using the denosing level and the sharpening level.
    Type: Grant
    Filed: May 13, 2020
    Date of Patent: May 10, 2022
    Assignee: AMBU A/S
    Inventors: Andreas Härstedt Jørgensen, Finn Sonnenborg
  • Patent number: 11321555
    Abstract: A method includes receiving a set of video frames that correspond to a video, including a first video frame and a second video frame that each include a face, wherein the second video frame is subsequent to the first video frame. The method further includes performing face tracking on the first video frame to identify a first face resampling keyframe and performing face tracking on the second video frame to identify a second face resampling keyframe. The method further includes deriving an interpolation amount. The method further includes determining a first interpolated face frame based on the first face resampling keyframe and the interpolation amount. The method further includes determining a second interpolated face frame based on the second face resampling keyframe and the interpolation amount. The method further includes rendering an interpolated first face and an interpolated second face. The method further includes displaying a final frame.
    Type: Grant
    Filed: November 11, 2019
    Date of Patent: May 3, 2022
    Assignee: Google LLC
    Inventor: Dillon Cower
  • Patent number: 11308590
    Abstract: This disclosure provides a system and method for generating an ultrasound image. The method includes receiving a raw ultrasound signal output from an ultrasound device and performing operations to transform the raw ultrasound signal into an enhanced ultrasound image. The enhanced ultrasound image may be further processed by radial symmetry filtering to generate a radial symmetry image. Both the enhanced image and the radial symmetry image can be by a medical practitioner to make a liver or kidney cancer diagnosis exclusively based on ultrasound data.
    Type: Grant
    Filed: February 27, 2018
    Date of Patent: April 19, 2022
    Assignee: RUTGERS, THE STATE UNIVERSITY OF NEW JERSEY
    Inventors: John Nosher, Ilker Hacihaliloglu
  • Patent number: 11301716
    Abstract: A method is provided for unsupervised domain adaptation for video classification. The method learns a transformation for each target video clips taken from a set of target videos, responsive to original features extracted from the target video clips. The transformation corrects differences between a target domain corresponding to target video clips and a source domain corresponding to source video clips taken from a set of source videos. The method adapts the target to the source domain by applying the transformation to the original features extracted to obtain transformed features for the plurality of target video clips. The method converts the original and transformed features of same ones of the target video clips into a single classification feature for each of the target videos. The method classifies a new target video relative to the set of source videos using the single classification feature for each of the target videos.
    Type: Grant
    Filed: July 18, 2019
    Date of Patent: April 12, 2022
    Inventors: Gaurav Sharma, Manmohan Chandraker, Jinwoo Choi
  • Patent number: 11301734
    Abstract: One embodiment provides a method, including: receiving, at an information handling device, user command input comprising identification of an object; receiving, at an information handling device, image input comprising at least one object; determining, using a processor, whether the object in the user command input is associated with the at least one object in the image input; and responsive to determining that the object in the user command input is associated with the at least one object, performing a function based on the determination.
    Type: Grant
    Filed: July 12, 2017
    Date of Patent: April 12, 2022
    Assignee: Lenovo (Singapore) Pte. Ltd.
    Inventors: Ming Qian, Song Wang, Jian Li
  • Patent number: 11301966
    Abstract: Various implementations disclosed herein include devices, systems, and methods for per-pixel filtering. In some implementations, a method includes obtaining an image data frame. In some implementations, the image data frame includes a plurality of pixels. In some implementations, the method includes generating a respective pixel characterization vector for each of the plurality of pixels. In some implementations, each pixel characterization vector includes an object label indicating an object type that the corresponding pixel of the plurality of pixels represents. In some implementations, the method includes modifying corresponding pixel data of the plurality of pixels having a first object label. In some implementations, the method includes synthesizing a first modified image data frame that includes modified pixel data for the plurality of pixels having the first object label and unmodified pixel data for the plurality of pixels not having the first object label.
    Type: Grant
    Filed: November 6, 2019
    Date of Patent: April 12, 2022
    Assignee: APPLE INC.
    Inventor: Jack Greasley
  • Patent number: 11295467
    Abstract: The present disclosure provides an object detection method, an object detection device, and a storage medium. In the present disclosure, with respect to each contour existing in an image to be detected, a contour diameter is calculated; a shape of the contour is determined based on side lengths of the circumscribed rectangular frame of each contour and the contour diameter; and an object in the image to be detected and parameter information of the object is determined based on the shape of each contour, thereby quickly and accurately detecting the shape of the object and reducing the labor cost and time of detection due to the low complexity of the detection.
    Type: Grant
    Filed: August 13, 2019
    Date of Patent: April 5, 2022
    Inventor: Mengmeng Wang
  • Patent number: 11294052
    Abstract: Described herein are methods and apparatus for increasing sensitivity of ultrasound imaging of fluid flow in an object of interest. Ultrasound imaging of blood perfusion can be performed without contrast enhancement. Embodiments include transforming a spatiotemporal echo data array into a three-dimensional perfusion data array having a spatial dimension, a slow-time dimension, and a frame-time dimension, and filtering the perfusion data array with an eigen passband clutter filter. The clutter filter can increase sensitivity and utility of ultrasound imaging of fluid flow. In some aspects, the method can yield blood flow signal power and perfusion values well separated from tissue clutter. In an example, enhancements to ischemic tissue perfusion maps in a murine model are shown.
    Type: Grant
    Filed: June 17, 2019
    Date of Patent: April 5, 2022
    Assignee: The Board of Trustees of the University of Illinois
    Inventors: Michael F. Insana, Minwoo Kim
  • Patent number: 11270111
    Abstract: Embodiments of the present invention are directed to a computer-implemented method for detecting vegetation near a power line. A non-limiting example of the computer-implemented method includes detecting the spectral characteristics of a target location. The spectral characteristics can be used to detect a tree at the target location. The location of the power lines that are within a buffer distance of the tree can be detected. Given that the tree is within a buffer distance, a risk value can be calculated for the tree. The risk value can convey the likelihood that the tree will contact the power line. If the risk is above a threshold amount, a work order can be issued to direct a crew to trim or cut the tree to prevent it from contacting the power line.
    Type: Grant
    Filed: February 4, 2019
    Date of Patent: March 8, 2022
    Assignee: INTERNATIONAL BUSINESS MACHINES CORPORATION
    Inventors: Levente I. Klein, John Wamburu, Hendrik F. Hamann, Bryan Sacks, Mahesh Sudhakaran
  • Patent number: 11256950
    Abstract: An image processing device converts an image that is a recognition object image to high-resolution, medium-resolution, and low-resolution images. The device sets the pixel of interest of the high-resolution image, and votes the co-occurrence in a gradient direction with offset pixels, the co-occurrence in the gradient direction pixels in the medium-resolution image, and the co-occurrence in the gradient direction pixels in the low-resolution image, to a co-occurrence matrix. The device creates such a co-occurrence matrix for each pixel combination and for each resolution. The device executes the process on each of the pixels of the high-resolution image, and creates a co-occurrence histogram wherein the elements of a plurality of co-occurrence matrices are arranged in a line. The device normalizes the co-occurrence histogram and extracts, as a feature quantity of the image, a vector quantity having as a component a frequency resulting from the normalization.
    Type: Grant
    Filed: January 31, 2018
    Date of Patent: February 22, 2022
    Assignees: AISIN CORPORATION, KYUSHU INSTITUTE OF TECHNOLOGY
    Inventors: Hideo Yamada, Kazuhiro Kuno, Masatoshi Shibata, Shuichi Enokida, Hakaru Tamukoh
  • Patent number: 11238568
    Abstract: A method and a system for reconstructing obstructed face portions are provided herein. The method may include the following steps: obtaining off-line 3D data, being 3D data of a head of a person not wearing a face-obstructing object, being an object which obstructs a portion of the face of the person; obtaining in real time, real-time 3D data, being 3D data of said head, wherein said person wears said face-obstructing object; applying a 3D transformation to at least a portion of the off-line 3D data, based on the real-time 3D data, to yield reconstructed real time 3D data, being real-time 3D data related to the obstructed face portions; and merging the reconstructed real time 3D data into the real-time 3D data. The system may implement the aforementioned steps over a computer processor.
    Type: Grant
    Filed: February 4, 2020
    Date of Patent: February 1, 2022
    Assignee: Facebook Technologies, LLC
    Inventors: Nadav Grossinger, Emil Alon
  • Patent number: 11238277
    Abstract: Systems and methods for processing documents based on a cardinal graph convolution network by generating cardinal graph representations representing words as single nodes with edges connected between neighbouring nodes in four cardinal directions. Features tensors are generated for nodes of the cardinal graph representation and the cardinal directions are encoded to generate an adjacency tensor having node neighbour indices. Entries of the adjacency tensor are transformed into a one-hot encoding of the node neighbour indices. Neighbourhood feature tensors are created over node indices and the features in each block may be scaled, convolved and reduced into new feature tensors.
    Type: Grant
    Filed: June 11, 2020
    Date of Patent: February 1, 2022
    Inventors: Rinon Gal, Roy Shilkrot, Amos Simantov
  • Patent number: 11222439
    Abstract: An image processing apparatus includes a first learner trained to detect an orientation of a face in an image including a human face for receiving an input of a target image including a human face and outputting first information associated with an orientation of a face included in the target image, at least one second learner trained, for the first information, to detect a position of a facial component in an image including a human face for receiving an input of the target image and outputting second information associated with a position of a facial component included in the target image, and a control unit that controls the first learner and the second learner. The control unit uses the first learner to obtain the first information from the target image, and uses the second learner corresponding to the first information to obtain the second information from the target image.
    Type: Grant
    Filed: August 20, 2019
    Date of Patent: January 11, 2022
    Assignee: OMRON Corporation
    Inventors: Tomohiro Yabuuchi, Koichi Kinoshita, Tomoyoshi Aizawa, Tadashi Hyuga, Hatsumi Aoi, Mei Uetani
  • Patent number: 11216685
    Abstract: A user device detects, in a field of view of the camera, a first side of a document, and determines first information associated with the first side of the document. The user device selects a first image resolution based on the first information and captures, by the camera, a first image of the first side of the document according to the first image resolution. The user device detects, in the field of view of the camera, a second side of the document, and determines second information associated with the second side of the document. The user device selects a second image resolution based on the second information, and captures, by the camera, a second image of the second side of the document according to the second image resolution. The user device performs an action related to the first image and the second image.
    Type: Grant
    Filed: October 24, 2019
    Date of Patent: January 4, 2022
    Assignee: Capital One Services, LLC
    Inventors: Jason Pribble, Daniel Alan Jarvis, Nicholas Capurso
  • Patent number: 11189009
    Abstract: The present invention is directed to providing an image processing apparatus and an image processing method capable of easily obtaining a combined image and a developed image in a case where a construction formed of a plurality of planes is divisionally imaged.
    Type: Grant
    Filed: August 28, 2019
    Date of Patent: November 30, 2021
    Assignee: FUJIFILM Corporation
    Inventor: Hiroaki Kikuchi
  • Patent number: 11176416
    Abstract: A perception device includes: a first neural network that performs a common process associated with perception of an object and thus outputs results of the common process; a second neural network that receives an output of the first neural network and outputs results of a first perception process of perceiving the characteristics of the object with a first accuracy; and a third neural network that receives the output of the first neural network and intermediate data which is generated by the second neural network in the course of the first perception process and outputs results of a second perception process of perceiving the characteristics of the object with a second accuracy which is higher than the first accuracy.
    Type: Grant
    Filed: April 23, 2018
    Date of Patent: November 16, 2021
    Assignee: TOYOTA JIDOSHA KABUSHIKI KAISHA
    Inventor: Daisuke Hashimoto
  • Patent number: 11164287
    Abstract: A LED controller includes a power distribution module and an interface to an external data bus. An image frame buffer is connected to the interface to receive image data. A separate logic module is connected to the interface and configured to modify image frame buffer output signals sent to an LED pixel array connected to the image frame buffer. The LED pixel array can project light according to a pattern and intensity defined at least in part by the image held in the image frame buffer.
    Type: Grant
    Filed: June 28, 2019
    Date of Patent: November 2, 2021
    Assignee: Lumileds LLC
    Inventors: Ronald Bonne, Jyoti Kiron Bhardwaj
  • Patent number: 11158040
    Abstract: In a method for identifying a robot arm responsible for creating a scratch on a wafer, at least one scratch mark on a wafer is detected. A first scratch dimension of the at least one scratch mark is determined. The determined first scratch dimension is compared to a plurality of first robot arm dimensions to generate a plurality of first comparing results, wherein the first comparing results respectively correspond to a plurality of robot arms. One of the robot arms is identified based on the first comparing results.
    Type: Grant
    Filed: June 5, 2019
    Date of Patent: October 26, 2021
    Assignee: TAIWAN SEMICONDUCTOR MANUFACTURING CO., LTD.
    Inventor: Yen-Liang Chen
  • Patent number: 11151356
    Abstract: Convolutional neural networks (CNNs) of a set of CNNs are evaluated using a test set of images (electron micrographs) associated with a selected particle type. A preferred CNN is selected based on the evaluation and used for processing electron micrographs of test samples. The test set of images can be obtained by manual selection or generated using a model of the selected particle type. Upon selection of images using the preferred CNN in processing additional electron micrographs, the selected images can be added to a training set or used as an additional training set to retrain the preferred CNN. In some examples, only selected layers of the preferred CNN are retrained. In other examples, two dimensional projections of based on particles of similar structure are used for CNN training or retraining.
    Type: Grant
    Filed: February 27, 2019
    Date of Patent: October 19, 2021
    Assignee: FEI Company
    Inventors: John Flanagan, Erik Franken, Maurice Peemen