Patents Examined by Andrae S Allison
  • Patent number: 10445949
    Abstract: Disclosed herein are systems and methods for measuring package dimensions. The systems and methods may include determining, by a computing device comprising a processor, pixel locations of a plurality of corners of a package within an image. The systems and methods may also include determining, by the computing device, the package dimensions based on the pixel locations.
    Type: Grant
    Filed: August 29, 2017
    Date of Patent: October 15, 2019
    Assignee: NCR Corporation
    Inventor: Nico Ryan James Geguiera Sy
  • Patent number: 10438097
    Abstract: According to an embodiment, a recognition device includes a candidate detection unit, a recognition unit, a matching unit, and a prohibition processing unit. The candidate detection unit detects, from an input image, character candidates each being a set of pixels estimated to include a character. The recognition unit recognizes each of the character candidates and generates one or more recognition candidates each being a character of a candidate as a recognition result. The matching unit matches each of the one or more recognition candidates with a knowledge dictionary in which a recognition target character string is modeled, and generates matching results obtained by matching a character string estimated to be included in the input image with the knowledge dictionary. The prohibition processing unit deletes, from the matching results, a matching result obtained by matching a character string including a prohibition target character string with the knowledge dictionary.
    Type: Grant
    Filed: September 20, 2017
    Date of Patent: October 8, 2019
    Assignees: KABUSHIKI KAISHA TOSHIBA, TOSHIBA DIGITAL SOLUTIONS CORPORATION
    Inventor: Tomohisa Suzuki
  • Patent number: 10430685
    Abstract: In one embodiment, a method includes receiving a plurality of input frames of a video sequence associated with a time t, training a convolutional network to predict one or more future frames of the video sequence from the plurality of input frames based on a generative model, and outputting a first future frame of the video sequence associated with a time t+1 as predicted by the generative model. The training may comprise using an adversarial model and an image gradient difference loss model. In addition, the training may comprise randomly selecting temporal sequences of a n×m grid of pixels from the plurality of input frames exhibiting a threshold of optical flow.
    Type: Grant
    Filed: November 16, 2017
    Date of Patent: October 1, 2019
    Assignee: Facebook, Inc.
    Inventors: Michael Fabien Mathieu, Camille Couprie, Yann Andre Le Cun
  • Patent number: 10430802
    Abstract: Role based assessment for an IT management system, includes maintaining a plurality of roles, each role attributable to a user type within an IT management system. Mappings are defined between the plurality of user roles and assets of the IT management system. An assessment for the IT management system is then assembled from the perspective of a selected one of the plurality of roles based on mappings between the selected user role and the assets.
    Type: Grant
    Filed: September 3, 2014
    Date of Patent: October 1, 2019
    Assignee: ENTIT SOFTWARE LLC
    Inventors: Jin Wang, Jun Lu, Lingyan Meng, Bing Gu, Keke Qi
  • Patent number: 10430636
    Abstract: A display panel includes an array substrate including organic light-emitting structures, and at least one fingerprint identification unit. Each organic light-emitting structure has a first side facing the array substrate, and the fingerprint identification unit is disposed in a display region on the first side of the organic light-emitting structure. The display panel also includes a plurality of spacers and a plurality of anti-crosstalk pads. When being projected onto the array substrate, the orthogonal projection of each spacer is disposed between orthogonal projections of adjacent organic light-emitting structures. Each anti-crosstalk pad is disposed between a corresponding spacer and the fingerprint identification unit. An orthogonal projection of the anti-crosstalk pad onto the array substrate include a first region overlapping with the orthogonal projection of the spacer onto the array substrate.
    Type: Grant
    Filed: February 7, 2018
    Date of Patent: October 1, 2019
    Assignee: Shanghai Tianma Micro-Electronics Co., Ltd.
    Inventors: Liang Xie, Yang Zeng, Lihua Wang, Haochi Yu
  • Patent number: 10423027
    Abstract: A display device panel includes a display area including pixels and a non-display area. Each pixel is connected to one of gate lines and one of data lines. The non-display area includes data pad sections. The non-display area further includes a depiction of a first information code and a depiction of a second information code. The depiction of the first information code is disposed between first two adjacent data pad sections and is apart from an outline of the non-display area by a first distance. The depiction of the second information code is disposed between second two adjacent data pad sections and is apart from the outline of the non-display area by a second distance different from the first distance.
    Type: Grant
    Filed: June 15, 2016
    Date of Patent: September 24, 2019
    Assignee: SAMSUNG DISPLAY CO., LTD.
    Inventors: Kang Young Lee, Sang Hyun Kang, Dae Hyun Kim, Seung Hwan Kim, Joo Yeol Lee, Jang Bog Ju
  • Patent number: 10417746
    Abstract: An image processing method includes the steps of: generating, based on data of a plurality of images acquired by performing imaging using a same image sensor, a weight coefficient corresponding to each image so that a weight coefficient corresponding to an image of which a brightness change is smooth is relatively large; integrating the data of the plurality of images using the weight coefficients to generate data of an integrated image; applying smoothing to the data of the integrated image to generate data of a smoothed integrated image; and generating, based on the data of the integrated image and the data of the smoothed integrated image, data of estimated fixed-pattern noise which is an estimated value of fixed-pattern noise attributable to the image sensor.
    Type: Grant
    Filed: November 17, 2016
    Date of Patent: September 17, 2019
    Assignee: Canon Kabushiki Kaisha
    Inventors: Tomochika Murakami, Kazuya Kodama
  • Patent number: 10408918
    Abstract: In a method for determining orientation of an object, raw image data of the sky is recorded using a sky polarimeter. One or more of Stokes parameters (S0, S1, S2), degree of linear polarization (DoLP), and angle of polarization (AoP) are calculated from the image data to produce a set of processed images. Last known position and time data of the object are obtained, and a known Sun azimuth and elevation are calculated using the last known position and time data. Roll and pitch of the object are found, and the roll and pitch data are used to find a zenith in the processed images. The yaw/heading of the object is determined using the difference between a polarization angle at the zenith and a calculated Sun azimuth.
    Type: Grant
    Filed: May 8, 2018
    Date of Patent: September 10, 2019
    Assignee: Polaris Sensor Technologies, Inc.
    Inventors: Todd Michael Aycock, David B. Chenault, Arthur Lompado, J. Larry Pezzaniti
  • Patent number: 10410053
    Abstract: A method for detecting an information card in an image is provided. The method includes performing a line detection to obtain two endpoints of a line segment corresponding to each of four sides of the information card; generating, a linear equation of the side; obtaining coordinates of four intersection points of the four sides of the information card; mapping the coordinates of the four intersection points to four corners of a rectangular box of the information card, to obtain a perspective transformation matrix; performing perspective transformation on image content encircled by four straight lines represented by the four linear equations to provide transformed image content; forming a gradient template according to a layout of information content on the information card; and using the gradient template to match with the transformed image content and determining whether the image content is a correct information card.
    Type: Grant
    Filed: September 26, 2017
    Date of Patent: September 10, 2019
    Assignee: TENCENT TECHNOLOGY (SHENZHEN) COMPANY LIMITED
    Inventors: Chengjie Wang, Hui Ni, Jilin Li
  • Patent number: 10402640
    Abstract: A method for schematizing fields in documents involves segmenting an image of a document into snippets, classifying a first, a second, and a third snippet of the snippets as a first label, a second label, and a data field, respectively. The method further includes generating, for a combination of the first label and the data field, a first image to be classified, and classifying the first image to be classified to obtain a first classification score. The method further includes generating, for a combination of the second label and the data field, a second image to be classified, and classifying the second image to be classified to obtain a second classification score. The method also includes determining that the first classification score is greater than the second classification score and identifying, based on the determining step, that the data field is associated with the first label.
    Type: Grant
    Filed: October 31, 2017
    Date of Patent: September 3, 2019
    Assignee: Intuit Inc.
    Inventors: Richard Becker, Kimia Hassanzadeh
  • Patent number: 10402986
    Abstract: In one embodiment, a method includes a computing system accessing a first training data comprising a first image and a second image and an associated optical flow estimation. The system may input (1) the first image into a first machine-learning model configured to generate a first output and (2) the optical flow estimation into a second machine-learning model configured to generate a second output. The first output of the first machine-learning model is associated with first image segments of a predetermined number, and the second output of the second machine-learning model is associated with transformations of the predetermined number. The first output, the transformations, and the first image are configured to generate an estimated image. The system trains the first machine-learning model and the second machine-learning model based on at least a comparison of the estimated image and the second image.
    Type: Grant
    Filed: December 20, 2017
    Date of Patent: September 3, 2019
    Assignee: Facebook, Inc.
    Inventors: Benjamin Ray, Anurag Ranjan
  • Patent number: 10402658
    Abstract: A video retrieval system is provided, that includes a set of servers, configured to retrieve a video sequence from a database and forward it to a requesting device responsive to a match between an input text and a caption for the video sequence. The servers are further configured to translate the video sequence into the caption by (A) applying a C3D to image frames of the video sequence to obtain therefor (i) intermediate feature representations across L convolutional layers and (ii) top-layer features, (B) producing a first word of the caption for the video sequence by applying the top-layer features to a LSTM, and (C) producing subsequent words of the caption by (i) dynamically performing spatiotemporal attention and layer attention using the representations to form a context vector, and (ii) applying the LSTM to the context vector, a previous word of the caption, and a hidden state of the LSTM.
    Type: Grant
    Filed: October 26, 2017
    Date of Patent: September 3, 2019
    Assignee: NEC Corporation
    Inventors: Renqiang Min, Yunchen Pu
  • Patent number: 10403000
    Abstract: A method for analyzing biological reaction systems is provided. The method includes receiving an image of a substrate including a plurality of reaction sites after a biological reaction has taken place. Next, the method includes removing a noise background from the first image. The method includes determining an initial position of each reaction site based on an intensity threshold to generate a initial position set, then refining the initial position set of each reaction site based on an expected pattern of locations of the plurality of reaction sites to generate a first refined position set. The method further includes determining a presence or absence of a fluorescent emission from each reaction site based on the first refined position set and the first image.
    Type: Grant
    Filed: February 19, 2018
    Date of Patent: September 3, 2019
    Assignee: Life Technologies Corporation
    Inventors: Yong Chu, Jeffrey Marks
  • Patent number: 10402997
    Abstract: This invention is to provide a technique capable of detecting a displacement of a positioning object with respect to a reference position in an actual space with high accuracy. A position of an imaging object detected in an imaged image is converted into a displacement of the positioning object with respect to the reference position in the actual space based on conversion information representing a correlative relationship between the position of the imaging object in the image and the displacement of the positioning object from the reference position. The conversion information is determined in advance based on positions of the imaging object detected from a plurality of images respectively imaged with the positioning object positioned at each of a plurality of imaging positions.
    Type: Grant
    Filed: October 25, 2016
    Date of Patent: September 3, 2019
    Assignee: SCREEN Holdings Co., Ltd.
    Inventors: Kazuhiro Kitamura, Hiroaki Kakuma
  • Patent number: 10395377
    Abstract: A method performed by an electronic device is described. The method includes generating a depth map of a scene external to a vehicle. The method also includes performing first processing in a first direction of a depth map to determine a first non-obstacle estimation of the scene. The method also includes performing second processing in a second direction of the depth map to determine a second non-obstacle estimation of the scene. The method further includes combining the first non-obstacle estimation and the second non-obstacle estimation to determine a non-obstacle map of the scene. The combining includes combining comprises selectively using a first reliability map of the first processing and/or a second reliability map of the second processing The method additionally includes navigating the vehicle using the non-obstacle map.
    Type: Grant
    Filed: January 16, 2018
    Date of Patent: August 27, 2019
    Assignee: QUALCOMM Incorporated
    Inventors: Gokce Dane, Vasudev Bhaskaran
  • Patent number: 10395131
    Abstract: In the case where a user extracts a desired character string by specifying a range by using a finger or the like of him/herself on an image including a character, a specific character (space or the like) located at a position adjacent to the desired character string is prevented from being included unintendedly in the selected range. The character area corresponding to each character included in the image is identified and character recognition processing is performed for each of the identified character areas. Then, from results of the character recognition processing, a specific character is determined and the character area corresponding to the determined specific character is extended. Then, the range selected by the user in the displayed image is acquired and character recognition results corresponding to a plurality of character areas included in the selected range are output.
    Type: Grant
    Filed: February 3, 2017
    Date of Patent: August 27, 2019
    Assignee: CANON KABUSHIKI KAISHA
    Inventor: Hiromasa Kawasaki
  • Patent number: 10387991
    Abstract: An apparatus and method for frame buffer compression. For example, one embodiment of a method comprises: identifying a wavelet function to compress an original frame buffer image; performing a wavelet transform on the original frame buffer image using the wavelet function; using a specified subset of wavelet coefficients of the wavelet function to construct an approximated image function; determining residuals comprising differences between the original frame buffer image and the approximated image function; and storing the residuals for subsequent use in recreating the original frame buffer image from the approximated image function.
    Type: Grant
    Filed: July 1, 2016
    Date of Patent: August 20, 2019
    Assignee: Intel Corporation
    Inventor: Tomas G. Akenine-Moller
  • Patent number: 10380461
    Abstract: Approaches introduce a pre-processing and post-processing framework to a neural network-based approach to identify items represented in an image. For example, a classifier that is trained on several categories can be provided. An image that includes a representation of an item of interest is obtained. Rotated versions of the image are generated and each of a subset of the rotated images is analyzed to determine a probability that a respective image includes an instance of a particular category. The probabilities can be used to determine a probability distribution of output category data, and the data can be analyzed to select an image of the rotated versions of the image. Thereafter, a categorization tree can then be utilized, whereby for the item of interest represented the image, the category of the item can be determined. The determined category can be provided to an item retrieval algorithm to determine primary content for the item of interest.
    Type: Grant
    Filed: October 20, 2017
    Date of Patent: August 13, 2019
    Assignee: A9.COM, INC.
    Inventors: Avinash Aghoram Ravichandran, Matias Omar Gregorio Benitez, Rahul Bhotika, Scott Daniel Helmer, Anshul Kumar Jain, Junxiong Jia, Rakesh Madhavan Nambiar, Oleg Rybakov
  • Patent number: 10366292
    Abstract: A system is provided for video captioning. The system includes a processor. The processor is configured to apply a three-dimensional Convolutional Neural Network (C3D) to image frames of a video sequence to obtain, for the video sequence, (i) intermediate feature representations across L convolutional layers and (ii) top-layer features. The processor is further configured to produce a first word of an output caption for the video sequence by applying the top-layer features to a Long Short Term Memory (LSTM). The processor is further configured to produce subsequent words of the output caption by (i) dynamically performing spatiotemporal attention and layer attention using the intermediate feature representations to form a context vector, and (ii) applying the LSTM to the context vector, a previous word of the output caption, and a hidden state of the LSTM. The system further includes a display device for displaying the output caption to a user.
    Type: Grant
    Filed: October 26, 2017
    Date of Patent: July 30, 2019
    Assignee: NEC Corporation
    Inventors: Renqiang Min, Yunchen Pu
  • Patent number: 10362321
    Abstract: In general, according to one embodiment, an image distribution device of the embodiment includes a receiver, a compressor, a request acceptor, a determiner, and a distributor. The receiver receives data captured by an imaging device of an in-vehicle device mounted in a vehicle. The compressor compresses the captured data of after elapse of a certain time to generate compressed captured data. The request acceptor accepts a distribution request to specify a desired time from an information processing device. The determiner determines whether the desired time is a past time within a certain time from a present time. The distributor distributes, when the desired time is the past time within the certain time from a present time, the captured data to the information processing device, and distributes, when the desired time is a past time prior to the certain time, the compressed captured data to the information processing device.
    Type: Grant
    Filed: August 22, 2017
    Date of Patent: July 23, 2019
    Assignees: Kabushiki Kaisha Toshiba, Toshiba Infrastructure Systems & Solutions Corporation
    Inventors: Hideki Ueno, Sunao Wada, Mikio Hamada, Junichi Nakamura