Artifact Removal Or Suppression (e.g., Distortion Correction) Patents (Class 382/275)
  • Patent number: 11164292
    Abstract: Provided are a system and method for correcting an image through estimation of a distortion parameter. The method includes receiving a distorted image including one or more measurement targets, extracting a plurality of feature points from each of the measurement targets, classifying the one or more measurement targets as a distorted target and an undistorted target by comparing distances between the plurality of extracted feature points and a center point of the received distorted image with each other, estimating a distortion parameter on the basis of standard deviations of a plurality of feature points of the classified distorted target and undistorted target, and correcting the received distorted image on the basis of the estimated distortion parameter.
    Type: Grant
    Filed: October 8, 2019
    Date of Patent: November 2, 2021
    Assignee: CHUNG ANG University industry Academic Cooperation Foundation
    Inventors: Joonki Paik, Minjung Lee, Hyungtae Kim
  • Patent number: 11158034
    Abstract: An image inspection device includes a mounting unit on which an image projection device that directly projects an image on a retina of a user is to be mounted; a condensing lens that condenses a light beam emitted from the image projection device mounted on the mounting unit; a detector on which an inspection image is projected and detected; and a controller that inspects the inspection image detected by the detector. The image inspection device inspects the image projected by the image projection device, which directly projects the image on the retina of the user.
    Type: Grant
    Filed: August 7, 2017
    Date of Patent: October 26, 2021
    Assignee: QD LASER, INC.
    Inventors: Seiji Morino, Makoto Suzuki, Manabu Ishimoto
  • Patent number: 11156689
    Abstract: The present disclosure related to techniques for extracting a respiratory navigation signal for use in a magnetic resonance imaging system. The extracted respiratory navigation signals accurately represent respiratory motions of a patient.
    Type: Grant
    Filed: April 17, 2020
    Date of Patent: October 26, 2021
    Assignee: Siemens Healthcare GmbH
    Inventors: Qiong Zhang, Wei Liu, Nan Xiao
  • Patent number: 11153588
    Abstract: An apparatus for video coding using dual deblocking filter thresholds may include a processor generating a reconstructed frame by decoding an encoded bitstream and outputting the reconstructed frame. Decoding may include generating a decoded block by decoding a portion of the encoded bitstream, identifying a first deblocking threshold index from the encoded bitstream, identifying a second deblocking threshold index from the encoded bitstream, generating a reconstructed block based on the decoded block, and including the reconstructed block in the reconstructed frame. Generating the reconstructed block may include deblocking based on the first deblocking threshold index and the second deblocking threshold index.
    Type: Grant
    Filed: April 2, 2020
    Date of Patent: October 19, 2021
    Assignee: GOOGLE LLC
    Inventors: Jingning Han, Yaowu Xu
  • Patent number: 11151702
    Abstract: Electronic devices, methods, and program storage devices for leveraging machine learning to perform improved image fusion and/or noise reduction are disclosed. An incoming image stream may be obtained from an image capture device, wherein the incoming image stream comprises a variety of differently-exposed captures, e.g., EV0 images, EV? images, EV+ images, long exposure images, EV0/EV? image pairs, etc., which are received according to a particular pattern. When a capture request is received, two or more intermediate assets may be generated based on determined combinations of images from the incoming image stream, and the intermediate assets may then be fed into a neural network that has been trained to determine one or more sets of parameters to optimally fuse and/or noise reduce the intermediate assets. In some embodiments, the network may be trained to operate on levels of pyramidal decompositions of the intermediate assets independently, for increased efficiency and memory utilization.
    Type: Grant
    Filed: September 9, 2019
    Date of Patent: October 19, 2021
    Assignee: Apple Inc.
    Inventors: Marius Tico, Tiffany Jou Cheng, Jun Hu
  • Patent number: 11144749
    Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for receiving mission specific data. Converting the mission specific data into attribute classes recognizable by image recognition classifiers, where the image recognition classifiers are pre-trained to detect objects corresponding to the attributed classes within digital images. Obtaining, from a wearable camera system, low resolution images of a scene and high resolution images of the scene. Detecting, within the low resolution images using a low resolution image classifier, an object that corresponds to one of the attribute classes. In response to detecting the object that corresponds to one of the attribute classes, providing, for presentation to a user of the wearable camera system by a presentation system, an first alert indicating a potential detection of the suspect, and providing the high resolution images to a high resolution image classifier.
    Type: Grant
    Filed: January 9, 2020
    Date of Patent: October 12, 2021
    Assignee: Idemia Identity & Security USA LLC
    Inventors: Peter Zhen-Ping Lo, Stephen Gregg, Thomas Gruschus
  • Patent number: 11140415
    Abstract: The present invention relates to a method for decoding a bitstream for a video signal and a device therefor, the method comprising the steps of: acquiring a value indicating the similarity between reference samples for intra prediction of a current block; determining an intra prediction mode of the current block on the basis of the acquired value and a predetermined threshold value; and reconstructing the current block on the basis of the determined intra prediction mode, wherein the step of determining the intra prediction mode of the current block comprises: when the value indicating the similarity is smaller than the predetermined threshold value, acquiring index information from the bitstream; preparing a list of candidate modes on the basis of intra prediction modes of blocks neighboring the current block, without acquiring flag information indicating whether the intra prediction mode of the current block is derived from the blocks neighboring the current block; and determining, as the intra prediction m
    Type: Grant
    Filed: November 30, 2018
    Date of Patent: October 5, 2021
    Assignee: LG ELECTRONICS INC.
    Inventors: Jin Heo, Seunghwan Kim, Jaehyun Lim
  • Patent number: 11107199
    Abstract: The disclosure analyzes moon-earth relationship and influencing factors for geometric distortion of a moon-based earth image, focuses on influences of position change of a sublunar point, a curvature of the earth, a terrain fluctuation on a large-scale hemisphere image, and the problem of small number of ground control points and uneven distribution thereof, and proposes a projection polar coordinate geometric expression method for the moon-based earth observation image, which takes into account of movement of the sublunar point, while considering multi-platform earth observation data, to realize accurate automatic geometric correction of the moon-based platform earth observation data.
    Type: Grant
    Filed: April 26, 2019
    Date of Patent: August 31, 2021
    Assignee: Institute of Remote Sensing and Digital Earth, Chinese Academy of Sciences
    Inventors: Huadong Guo, Lu Zhang, Guang Liu
  • Patent number: 11100352
    Abstract: Disclosed is a computer-readable medium including a program code that, when executed by processing circuitry, causes the processing circuitry to generate a feature map from an input image, to extract a region of interest from the feature map, and to generate a predicted mask based on the region of interest. The processing circuitry may use a predicted mask and a real mask to learn a convolutional neural network system. The real mask includes first pixels corresponding to the real boundary and second pixels corresponding to a fake boundary adjacent to the real boundary.
    Type: Grant
    Filed: September 19, 2019
    Date of Patent: August 24, 2021
    Assignees: SAMSUNG ELECTRONICS CO., LTD., AJOU UNIVERSITY INDUSTRY-ACADEMIC COOPERATION FOUNDATION
    Inventors: Hyunku Lee, Hyunsurk Ryu, Ha Young Kim, BaRom Kang
  • Patent number: 11100653
    Abstract: An image recognition apparatus is provided with: a divider configured to divide a frame image into a plurality of areas, wherein the frame image is obtained by imaging or capturing a scene outside through a window glass on which a wiper is mounted, from an interior of a moving body; a determinator configured to determine whether or not at least one of the wiper and an attached substance to the window glass appears, for each of the plurality of areas; an object detector configured to detect an object included in such an area of the plurality of areas that it is determined that the wiper and the attached substance do not appear; and a tracking device configured to track a tracking target, which is an object included in two or more frame images, on the basis of a detection result of the object detector.
    Type: Grant
    Filed: July 17, 2019
    Date of Patent: August 24, 2021
    Assignees: TOYOTA JIDOSHA KABUSHIKI KAISHA, DENSO CORPORATION
    Inventors: Masataka Yokota, Jia Sun, Takahiro Sota
  • Patent number: 11087681
    Abstract: A display device includes a display panel including a first pixel that emits light with a first luminance that is lower than target luminance and a second pixel that emits light with a second luminance that is higher than the target luminance, a sensor configured to measure a first characteristic of a first light emitting element in the first pixel and a second characteristic of a second light emitting element in the second pixel, and a data compensator configured to calculate a degradation amount of the second pixel based on the first characteristic and the second characteristic.
    Type: Grant
    Filed: January 15, 2020
    Date of Patent: August 10, 2021
    Assignee: Samsung Display Co., Ltd.
    Inventor: Jin-Woo Park
  • Patent number: 11074677
    Abstract: Methods for recovering saturated pixel values in raw pixel data are described. Given an image with raw pixel values captured using sensors with a color filter array (CFA), image regions are classified according to how many color channels of the CFA are saturated. Values of saturated pixels are estimated based on weighted color ratios of unsaturated pixels and recovered pixels. Example methods for a Bayer CFA are provided.
    Type: Grant
    Filed: March 14, 2019
    Date of Patent: July 27, 2021
    Assignee: Dolby Laboratories Licensing Corporation
    Inventors: Jon S. McElvain, Robin Atkins
  • Patent number: 11049225
    Abstract: A video processing system includes: an acquirer that acquires video data including a main video; a format acquirer that acquires a display format that indicates a luminance dynamic range format displayable by a video display apparatus; a generator that generates first characteristics information that indicates first dynamic luminance characteristics that correspond to the display format by using the video data when a luminance dynamic range format of the main video indicated by the video data is different from the display format, the first dynamic luminance characteristics being dynamic luminance characteristics indicating a time-dependent change in luminance characteristics of the main video; and a video transmitter that outputs the first characteristics information generated by the generator.
    Type: Grant
    Filed: February 26, 2018
    Date of Patent: June 29, 2021
    Assignee: PANASONIC INTELLECTUAL PROPERTY MANAGEMENT CO., LTD.
    Inventors: Masayuki Kozuka, Masaya Yamamoto, Toshiroh Nishio, Kazuhiko Kouno, Yoshiichiro Kashiwagi, Takeshi Hirota, Hiroshi Yahata, Yoshihiro Mori
  • Patent number: 11042972
    Abstract: An image processing apparatus includes a receiving unit that receives a first image that is captured with an object irradiated with light from a light irradiating unit and a second image that is captured with the object not irradiated with light from the light irradiating unit, and a generating unit that generates a composite image by adjusting brightness of a whole area of the second image such that brightness at a specific location in the first image is approximately equal to brightness at a corresponding specific location in the second image, and by replacing in the first image a pixel value of a pixel that is affected by light irradiation by the light irradiating unit with a pixel value of a pixel in the second image that has been adjusted in brightness.
    Type: Grant
    Filed: February 27, 2019
    Date of Patent: June 22, 2021
    Assignee: FUJIFILM Business Innovation Corp.
    Inventor: Shinya Miyamori
  • Patent number: 11042724
    Abstract: An image processing device image processing device (100), in which correction is less likely to be impossible or correction is less likely to be inappropriate as compared with that in the related art, performs, on an image including a face (500), correction processing to reduce/expand a first correction region (501) of a circle that is located at a position corresponding to a center of the face (500) and that has a size depending on a size of the face (500) and to expand/reduce a second correction region (502) around the first correction region.
    Type: Grant
    Filed: November 29, 2017
    Date of Patent: June 22, 2021
    Assignee: SHARP KABUSHIKI KAISHA
    Inventors: Keisuke Omori, Kei Tokui
  • Patent number: 11037274
    Abstract: Supervised machine learning using neural networks is applied to denoising images rendered by MC path tracing. Specialization of neural networks may be achieved by using a modular design that allows reusing trained components in different networks and facilitates easy debugging and incremental building of complex structures. Specialization may also be achieved by using progressive neural networks. In some embodiments, training of a neural-network based denoiser may use importance sampling, where more challenging patches or patches including areas of particular interests within a training dataset are selected with higher probabilities than others. In some other embodiments, generative adversarial networks (GANs) may be used for training a machine-learning based denoiser as an alternative to using pre-defined loss functions.
    Type: Grant
    Filed: February 12, 2020
    Date of Patent: June 15, 2021
    Assignees: Pixar, Disney Enterprises, Inc.
    Inventors: Thijs Vogels, Fabrice Rousselle, Brian McWilliams, Mark Meyer, Jan Novak
  • Patent number: 11030502
    Abstract: An omnidirectional barcode may include a graphical representation of a one-dimensional barcode wherein the graphical representation of the one-dimensional barcode is extended circumferentially around a three-dimensional (3D) object. A method of forming an omnidirectional barcode may include creating a graphical representation of a one-dimensional barcode and extending the one-dimensional representation in at least one dimension.
    Type: Grant
    Filed: February 6, 2017
    Date of Patent: June 8, 2021
    Assignee: Hewlett-Packard Development Company, L.P.
    Inventors: Robert Ulichney, Matthew D. Gaubatz
  • Patent number: 11032476
    Abstract: An image sensor includes a pixel array that includes a plurality of pixels, a first interface directly connected to an external gyro sensor and that receives gyro data output by the gyro sensor in response to motion, and a control logic that generates image data by exposing the plurality of pixels for a predetermined exposure period, generates valid data that correspond to the exposure period using the gyro data, and generates, based on the valid data, compensation information that represents a movement path of the motion.
    Type: Grant
    Filed: August 12, 2019
    Date of Patent: June 8, 2021
    Assignee: SAMSUNG ELECTRONICS CO., LTD.
    Inventors: Yong Seong Kim, Da Hee Lee, Young Mi Jo, Dong Ki Min, Joon Hyuk Im
  • Patent number: 11025947
    Abstract: A motion vector (MV) projection method includes generating motion field motion vectors (MFMVs) for a first portion of a current frame by applying MV projection to MVs of a portion of each of reference frames and storing the MFMVs of the first portion of the current frame into an MFMV buffer, and generating MFMVs for a second portion of the current frame by applying MV projection to MVs of a portion of each of the reference frames and storing the MFMVs of the second portion of the current frame into the MFMV buffer. The second portion does not overlap the first portion. Before generating the MFMVs for the second portion of the current frame is done, at least one of the MFMVs of the first portion is read from the MFMV buffer and involved in motion vector determination of at least one coding block included in the first portion.
    Type: Grant
    Filed: November 26, 2019
    Date of Patent: June 1, 2021
    Assignee: MEDIATEK INC.
    Inventors: Yung-Chang Chang, Chia-Yun Cheng, Cheng-Han Li, Hong-Cheng Lin, Chi-Hung Chen
  • Patent number: 11017504
    Abstract: A system includes an image sensor, an imaging pipeline, and a display device. The image sensor is configured to capture a first frame of pixel data. The imaging pipeline is coupled to the image sensor to receive the first frame of pixel data. The imaging pipeline includes an adaptive noise filter. The adaptive noise filter is configured to filter a pixel based on noise in the pixel. The imaging pipeline is configured to output a second frame of pixel data. The second frame of pixel data includes pixels filtered by the adaptive noise filter. The display device is coupled to the imaging pipeline to receive the second frame of pixel data. The display device being configured to display the second frame of pixel data.
    Type: Grant
    Filed: January 27, 2017
    Date of Patent: May 25, 2021
    Assignee: Intuitive Surgical Operations, Inc.
    Inventors: Max J. Trejo, Jeffrey M. DiCarlo
  • Patent number: 11009446
    Abstract: Reader observation and monitoring of the imaging of biological development is shown and described. In one embodiment, a method for monitoring biological growth includes identifying an alert from comparing a background value identified on a plate to a non-background value identified on a plate. In particular examples, a pre-filter, threshold, and histogram analysis identifies an alert count.
    Type: Grant
    Filed: October 29, 2019
    Date of Patent: May 18, 2021
    Assignee: Charm Sciences, Inc.
    Inventors: Richard T. Skiffington, Robert S. Salter, Robert J. Markovsky, Stanley E. Charm, Paul E. Graham, Jovo Djuragic
  • Patent number: 11006095
    Abstract: Various embodiments of the present disclosure relate generally to systems and methods for drone-based systems and methods for capturing a multi-media representation of an entity. In some embodiments, the multi-media representation is digital, or multi-view, or interactive, and/or the combinations thereof. According to particular embodiments, a drone having a camera is controlled or operated to obtain a plurality of images having location information. The plurality of images, including at least a portion of overlapping subject matter, are fused to form multi-view interactive digital media representations.
    Type: Grant
    Filed: March 5, 2018
    Date of Patent: May 11, 2021
    Assignee: Fyusion, Inc.
    Inventors: Stefan Johannes Josef Holzer, Stephen David Miller, Radu Bogdan Rusu
  • Patent number: 10989821
    Abstract: A coded aperture mask is provided. The coded aperture mask may include a 2-D planar substrate having a plurality of holes constructed based on a Cartesian product of a first 1-D aperture set and a second 1-D aperture set. The first 1-D aperture set may have a first balanced decoder. The second 1-D aperture set may have a second balanced decoder. The Cartesian product may involve the first 1-D aperture set and the second 1-D aperture set arranged in a non-zero angle (e.g., 90 degrees) to each other. The first 1-D aperture set may define a first axis of the 2-D planar substrate. The second 1-D aperture set may define a second axis of the 2-D planar substrate. The plurality of holes on the 2-D planar substrate may correspond to holes in both of the first 1-D aperture set and the second 1-D aperture set.
    Type: Grant
    Filed: December 14, 2017
    Date of Patent: April 27, 2021
    Assignees: Nanyang Technological University, Auckland University of Technology
    Inventors: Paul M. E. Shutler, Kevin Byard
  • Patent number: 10981376
    Abstract: An inspection apparatus that inspects printed matter generated based on print data by using a read image of a printed surface of the printed matter includes processing circuitry. The processing circuitry performs acquisition, calculation, correction, and comparison. The processing circuitry acquires the read image as an inspection target image, and acquires a reference image based on the print data. The processing circuitry calculates a flare inverse correction value based on flare that occurs at time of creation of the inspection target image. The processing circuitry performs flare inverse correction on the reference image based on the flare inverse correction value to acquire a corrected reference image. The processing circuitry compares the corrected reference image with the inspection target image to perform inspection.
    Type: Grant
    Filed: June 27, 2019
    Date of Patent: April 20, 2021
    Assignee: RICOH COMPANY, LTD.
    Inventor: Takuji Kamada
  • Patent number: 10972729
    Abstract: Codecs that use larger blocks may have larger boundary regions that may benefit from filtering. In some embodiments, the deblocking filter determines filters and/or a number of samples of the block to be filtered based on block dimensions. For example, In one embodiment, deblocking filter parameters for the video block are determined based on at least one dimension of the size of the video block. The filter parameters include a filter to be applied or a number of pixels along a boundary with a neighboring block to which the filter is to be applied determined based on the at least one dimension.
    Type: Grant
    Filed: January 17, 2019
    Date of Patent: April 6, 2021
    Assignee: QUALCOMM Incorporated
    Inventors: Akshay Gadde, Li Zhang, Wei-Jung Chien, Marta Karczewicz
  • Patent number: 10959665
    Abstract: Devices and methods for determining the thickness of hair are provided. An exemplary device includes a computing unit with a processor and a local memory. The processor is connected to the local memory and configured to read data from and/or to write data into the local memory. The device further includes a user input/output unit configured for interaction with a user. Also, the computing unit includes an image data input interface for receiving image data of at least one individual hair. The image data input interface is connected to the processor. Further, the computing unit includes a user interface connected to the user input/output unit and to the processor. The computing unit is configured to process the image data of the at least one individual hair to create hair thickness values of the at least one individual hair and to output the hair thickness values to the user interface.
    Type: Grant
    Filed: December 13, 2018
    Date of Patent: March 30, 2021
    Assignee: Henkel AG & Co. KGaA
    Inventors: Georg Knuebel, Lucile Bonnin, Andreas Kirch
  • Patent number: 10949988
    Abstract: An information processing apparatus includes an acquiring unit configured to acquire first data and second data, a deriving unit configured to derive gradient data describing a gradient of correspondence data on the basis of the correspondence data describing a correspondence between partial data of the first data and partial data of the second data, and a propagating unit configured to propagate the correspondence data on the basis of the gradient data.
    Type: Grant
    Filed: April 7, 2017
    Date of Patent: March 16, 2021
    Assignee: Canon Kabushiki Kaisha
    Inventors: Hisayoshi Furihata, Masahiro Suzuki, Akihiro Katayama, Masakazu Fujiki
  • Patent number: 10949951
    Abstract: Systems and methods for improved image denoising using a deep learning network model are disclosed. An example system includes an input data processor to process a first patient image of a first patient to add a first noise to the first patient image to form a noisy image input. The example system includes an image data denoiser to process the noisy image input using a first deep learning network to identify the first noise. The image data denoiser is to train the first deep learning network using the noisy image input. When the first deep learning network is trained to identify the first noise, the image data denoiser is to deploy the first deep learning network as a second deep learning network model to be applied to a second patient image of the first patient to identify a second noise in the second patient image.
    Type: Grant
    Filed: August 23, 2018
    Date of Patent: March 16, 2021
    Assignee: General Electric Company
    Inventors: Jie Tang, Eric Gros, Jiang Hsieh, Roy Nilsen
  • Patent number: 10930185
    Abstract: An information processing apparatus provided with a display portion including a plurality of sub-pixels every pixel, connected to an output apparatus for presenting an image displayed on the display portion to a user through a lens, and producing the image to be displayed on the display portion receives source image data becoming an object of display, obtains a corresponding pixel on the source image data by information on a position of the sub-pixel and a predetermined conversion expression every sub-pixel included in a pixel with respect to the respective pixels constituting corrected image data, and determines luminances of the respective sub-pixels based on a pixel value of the pixel on the source image obtained every sub-pixel, thereby producing the corrected image data.
    Type: Grant
    Filed: May 19, 2016
    Date of Patent: February 23, 2021
    Assignee: Sony Interactive Entertainment Inc.
    Inventors: Atsushi Takizawa, Takamasa Araki, Rei Miyazaki, Shotaro Tada
  • Patent number: 10909668
    Abstract: Embodiments are disclosed that apply adaptive sub-tiles to captured images for distortion correction in vision-based assistance systems and methods. A captured image is processed to generate corrected tiles, and selected numbers of sub-tiles are used to generate each of the corrected tiles depending upon the pixel densities for regions of the captured image. The corrected sub-tiles are combined to form corrected tiles, and corrected tiles are combined for form a corrected image. The corrected image can be used to output control signals to cause actions to be issued to a user of the system such as a driver of a vehicle. For one embodiment, the corrected tiles are generated one at a time, and corrected sub-tiles for each corrected tile are also generated one at a time based upon individual source data blocks determined by a pre-determined sub-tile configuration. Efficient memory use and data transfers are provided.
    Type: Grant
    Filed: July 31, 2019
    Date of Patent: February 2, 2021
    Assignee: NXP USA, Inc.
    Inventors: Sharath Subramanya Naidu, Michael Andreas Staudenmaier, Chanpreet Singh, Rahul Jain
  • Patent number: 10902562
    Abstract: A system for filtering an input image dataset is disclosed. A plurality of sequences of input signal values, wherein each sequence corresponds to a different attribute, wherein an input signal value is associated with an attribute and a sampling point i. At least one processor 4 is configured to control computing an output signal value corresponding to a sampling point and an attribute. For a particular sampling point i and for each of a plurality of different attributes of the set of attributes, associating a weight to an input signal value based on a similarity between the signal value and for a plurality of the sampling points j excluding the sampling point i. Also the system computes a weighted sum based on the input signal values and weights. The attribute is a location or a frequency.
    Type: Grant
    Filed: May 15, 2017
    Date of Patent: January 26, 2021
    Assignee: Stichting Katholieke Universiteit
    Inventors: Ewoud Joris Smit, Wolfgang Mathias Prokop
  • Patent number: 10896486
    Abstract: Embodiments and processes of computer tomography perform tasks associated with denoising a reconstructed image using an anisotropic diffusion filter and adaptively weighting an iterative instance of the diffused image based upon the product of a weight value and a difference between the iterative instance of the diffused image and the original image. In general, the adaptive weighting is a negative feedback in the iterative steps.
    Type: Grant
    Filed: December 11, 2014
    Date of Patent: January 19, 2021
    Assignee: TOSHIBA MEDICAL SYSTEMS CORPORATION
    Inventors: Zhi Yang, Michael D. Silver
  • Patent number: 10872637
    Abstract: Certain aspects involve video inpainting in which content is propagated from a user-provided reference frame to other video frames depicting a scene. For example, a computing system accesses a set of video frames with annotations identifying a target region to be modified. The computing system determines a motion of the target region's boundary across the set of video frames, and also interpolates pixel motion within the target region across the set of video frames. The computing system also inserts, responsive to user input, a reference frame into the set of video frames. The reference frame can include reference color data from a user-specified modification to the target region. The computing system can use the reference color data and the interpolated motion to update color data in the target region across set of video frames.
    Type: Grant
    Filed: September 27, 2019
    Date of Patent: December 22, 2020
    Assignee: Adobe Inc.
    Inventors: Geoffrey Oxholm, Seth Walker, Ramiz Sheikh, Oliver Wang, John Nelson
  • Patent number: 10867392
    Abstract: Methods and apparatus for generating improved image data from received input image data comprising first input image data associated with a first exposure level and second input image data associated with a second, different, exposure level. Motion detection data is generated from the received input image data by applying a noise model and improved image data is generated by combining data from the first and second input data in dependence on the motion detection data.
    Type: Grant
    Filed: October 3, 2018
    Date of Patent: December 15, 2020
    Assignee: Apical Limited
    Inventors: David Hanwell, Alassane Seck, Alex Kornienko
  • Patent number: 10855956
    Abstract: A system and a method are provided that prevent a wrong panel of a structure such as an aircraft from being opened for inspection of an interior of the structure. A system includes a computer, a portable terminal and a camera configured to capture an image of a structure with panels attached to generate a first image and capture another image of the structure with a predetermined panel among the panels removed to generate a second image and configured to send the first image and the second image to the computer. The computer combines the first image and the second image to generate a third image and sends the third image to the portable terminal. The portable terminal displays the third image. In the third image, the predetermined panel is translucently shown with at least one component at an interior of the predetermined panel being seen through.
    Type: Grant
    Filed: September 6, 2019
    Date of Patent: December 1, 2020
    Assignees: JAPAN AIRLINES CO., LTD., JAL ENGINEERING CO., LTD.
    Inventor: Naofumi Ito
  • Patent number: 10846828
    Abstract: The present disclosure relates to using a neural network to efficiently denoise images that were generated by a ray tracer. The neural network can be trained using noisy images generated with noisy samples and corresponding denoised or high-sampled images (e.g., many random samples). An input feature to the neural network can include color from pixels of an image. Other input features to the neural network, which would not be known in normal image processing, can include shading normal, depth, albedo, and other characteristics available from a computer-generated scene. After the neural network is trained, a noisy image that the neural network has not seen before can have noise removed without needing manual intervention.
    Type: Grant
    Filed: April 22, 2019
    Date of Patent: November 24, 2020
    Assignee: Pixar
    Inventors: Mark Meyer, Anthony DeRose, Steve Bako
  • Patent number: 10846863
    Abstract: There is provided an image processing apparatus in which a part used in predetermined processing is specified with use of a relatively small binarization threshold, from parts that have been converted into black pixels through binarization processing using a relatively large binarization threshold.
    Type: Grant
    Filed: May 14, 2018
    Date of Patent: November 24, 2020
    Assignee: CANON KABUSHIKI KAISHA
    Inventor: Takashi Miyauchi
  • Patent number: 10841521
    Abstract: The present disclosure relates to an information processing device, an information processing method, and a program that enable generation of stable PUFs. The information processing device includes: a reading unit that reads output data from a predetermined element a plurality of times; an average value calculation unit that calculates average values of the output data read by the reading unit; a median calculation unit that calculates the median of the average values calculated by the average value calculation unit; and a PUF generation unit that generates a physical unclonable function (PUF) by comparing the median with the average values. The predetermined element is an image sensor, and the reading unit reads output data from the image sensor when the image sensor is shielded from light. The present technology can be applied to imaging devices, for example.
    Type: Grant
    Filed: December 8, 2016
    Date of Patent: November 17, 2020
    Assignee: Sony Corporation
    Inventors: Koichi Baba, Taiichiro Watanabe, Hirokazu Ejiri
  • Patent number: 10832074
    Abstract: From a first image using a model, a first uncertainty map is generated. An uncertainty level of a location in the first uncertainty map corresponds to a detection of a known structure in a portion of the first image. A first weighted image corresponding to the first uncertainty map is generated, the generating including assigning a first weight to a pixel of the first image, the first weight corresponding to the uncertainty level of a location in the first uncertainty map corresponding to the pixel. From a second image using a model, a second uncertainty map is generated. A second weighted image corresponding to the second uncertainty map is generated. The first image and the second image are combined to form a composite image, each image participating in the composite image according to the corresponding weighted image.
    Type: Grant
    Filed: March 8, 2019
    Date of Patent: November 10, 2020
    Assignee: INTERNATIONAL BUSINESS MACHINES CORPORATION
    Inventors: Suman Sedai, Bhavna Josephine Antony, Kerry Halupka, Dwarikanath Mahapatra, Rahil Garnavi
  • Patent number: 10832381
    Abstract: A method for processing at least one X-ray image is provided. A variance of noise is signal dependent. The method includes applying a variance-stabilizing transformation to image data of the at least X-ray image to generate variance-stabilized data. At least one transform parameter of the variance-stabilizing transformation is dependent on a property of the at least one X-ray image that depends on an X-ray imaging device and/or a measurement parameter used to record the at least one X-ray image. A noise reduction algorithm is applied to the variance-stabilized data to generate noise-reduced data, and an inverse transformation of the variance stabilizing transformation is applied to the noise-reduced data to generate a denoised X-ray image.
    Type: Grant
    Filed: July 27, 2018
    Date of Patent: November 10, 2020
    Assignee: Siemens Healthcare GmbH
    Inventors: Norbert Strobel, Sai Gokul Hariharan
  • Patent number: 10825426
    Abstract: A method of generating a high dynamic range (HDR) image is provided that includes capturing a long exposure image and a short exposure image of a scene, computing a merging weight for each pixel location of the long exposure image based on a pixel value of the pixel location and a saturation threshold, and computing a pixel value for each pixel location of the HDR image as a weighted sum of corresponding pixel values in the long exposure image and the short exposure image, wherein a weight applied to a pixel value of the pixel location of the short exposure image and a weight applied to a pixel value of the pixel location in the pixel long exposure image are determined based on the merging weight computed for the pixel location and responsive to motion in a scene of the long exposure image and the short exposure image.
    Type: Grant
    Filed: February 12, 2019
    Date of Patent: November 3, 2020
    Assignee: TEXAS INSTRUMENTS INCORPORATED
    Inventors: Rajesh Narasimha, Aziz Umit Batur
  • Patent number: 10825140
    Abstract: Provided are methods and apparatus related to Artificial Intelligence (AI) downscaling and upscaling and techniques related to reducing artifact problems. Some embodiments include down-scaling an original image through a Deep Neural Network (DNN); generating, from the original image and based on frequency transform coefficients, artifact information representing a region in the first image including an artifact in the first image. Post-processing may be performed based on the artifact information to change pixels in the first image, thus reducing the effect of artifacts.
    Type: Grant
    Filed: March 23, 2020
    Date of Patent: November 3, 2020
    Assignee: SAMSUNG ELECTRONICS CO., LTD.
    Inventors: Heechul Yang, Jongseok Lee, Youngo Park, Kwangpyo Choi
  • Patent number: 10817990
    Abstract: Provided are methods and apparatus related to Artificial Intelligence (AI) downscaling and upscaling and techniques related to reducing artifact problems. Some embodiments include down-scaling an original image through a Deep Neural Network (DNN); generating, from the original image and based on frequency transform coefficients, artifact information representing a region in the first image including an artifact in the first image. Post-processing may be performed based on the artifact information to change pixels in the first image, thus reducing the effect of artifacts.
    Type: Grant
    Filed: April 16, 2020
    Date of Patent: October 27, 2020
    Assignee: SAMSUNG ELECTRONICS CO., LTD.
    Inventors: Heechul Yang, Jongseok Lee, Youngo Park, Kwangpyo Choi
  • Patent number: 10818026
    Abstract: Systems and methods for hybrid depth regularization in accordance with various embodiments of the invention are disclosed. In one embodiment of the invention, a depth sensing system comprises a plurality of cameras; a processor; and a memory containing an image processing application. The image processing application may direct the processor to obtain image data for a plurality of images from multiple viewpoints, the image data comprising a reference image and at least one alternate view image; generate a raw depth map using a first depth estimation process, and a confidence map; and generate a regularized depth map. The regularized depth map may be generated by computing a secondary depth map using a second different depth estimation process; and computing a composite depth map by selecting depth estimates from the raw depth map and the secondary depth map based on the confidence map.
    Type: Grant
    Filed: November 15, 2019
    Date of Patent: October 27, 2020
    Assignee: FotoNation Limited
    Inventors: Ankit Jain, Priyam Chatterjee, Kartik Venkataraman
  • Patent number: 10789686
    Abstract: Supervised machine learning using neural networks is applied to denoising images rendered by MC path tracing. Specialization of neural networks may be achieved by using a modular design that allows reusing trained components in different networks and facilitates easy debugging and incremental building of complex structures. Specialization may also be achieved by using progressive neural networks. In some embodiments, training of a neural-network based denoiser may use importance sampling, where more challenging patches or patches including areas of particular interests within a training dataset are selected with higher probabilities than others. In some other embodiments, generative adversarial networks (GANs) may be used for training a machine-learning based denoiser as an alternative to using pre-defined loss functions.
    Type: Grant
    Filed: January 6, 2020
    Date of Patent: September 29, 2020
    Assignees: Pixar, Disney Enterprises, Inc.
    Inventors: Thijs Vogels, Fabrice Rousselle, Brian McWilliams, Mark Meyer, Jan Novak
  • Patent number: 10776978
    Abstract: A method is provided for the automated identification of real world objects. The invention, implementable via various means such as a processing system, method, or data structure in a recording medium such as memory or as a self-contained electronic circuit, has wide ranging applicability to numerous fields such as in a user interface for gaming systems, like Kinect, or immersive environments such as remote surgical operations, and other medical diagnostic applications. Similarly, our method can be utilized in artificial intelligence application for automated robotic identification of targets, such as drone assisted search and rescue missions, and the enablement of patent protection against unlawful replication on 3D printers.
    Type: Grant
    Filed: April 8, 2013
    Date of Patent: September 15, 2020
    Inventor: Jeffrey W. Holcomb
  • Patent number: 10778885
    Abstract: A method and system for detecting facial expressions in digital images and applications therefore are disclosed. Analysis of a digital image determines whether or not a smile and/or blink is present on a person's face. Face recognition, and/or a pose or illumination condition determination, permits application of a specific, relatively small classifier cascade.
    Type: Grant
    Filed: April 9, 2018
    Date of Patent: September 15, 2020
    Assignee: FotoNation Limited
    Inventors: Catalina Neghina, Mihnea Gangea, Stefan Petrescu, Emilian David, Petronel Bigioi, Eric Zarakov, Eran Steinberg
  • Patent number: 10776909
    Abstract: The disclosure includes: a memory section, storing a learned model and an internal parameter set in the learned model; an acquisition section, acquiring an image of an inspection object photographed under predetermined conditions; a preprocessing section, generating a predetermined preprocessing filter according to a feature value in a preprocessing image being an image of the inspection object that is acquired by the acquisition section and that includes a defect, and a feature value corresponding to the internal parameter, and generating a preprocessed image by applying the generated preprocessing filter to an inspection image being an image of the inspection object that is acquired by the acquisition section and converting the inspection image; and an inspection section, inspecting the preprocessed image for presence or absence of the defect of the inspection object by using the stored learned model.
    Type: Grant
    Filed: September 13, 2018
    Date of Patent: September 15, 2020
    Assignee: OMRON Corporation
    Inventors: Yasuyuki Ikeda, Masashi Kurita
  • Patent number: 10766342
    Abstract: A shading device comprises: a shading member formed using a dimming glass plate capable of changing light transmittance, the shading member having a plate shape; a display apparatus being capable of transmitting light and disposed on a surface of the shading member, the surface being to face an operator during use of the shading member, in such a manner that a display portion faces the operator; an image pickup device to pick up, as an image, a region which an opposite surface of the surface faces, and generate image pickup data; a data processing circuit to generate display image data to be displayed on the display portion during use of the shading member, based on the image pickup data generated by the image pickup device; and a switch to change light transmittance of the dimming glass plate.
    Type: Grant
    Filed: April 14, 2017
    Date of Patent: September 8, 2020
    Assignee: Sakai Display Products Corporation
    Inventor: Katsuhiko Kishimoto
  • Patent number: 10755391
    Abstract: Digital image completion by learning generation and patch matching jointly is described. Initially, a digital image having at least one hole is received. This holey digital image is provided as input to an image completer formed with a dual-stage framework that combines a coarse image neural network and an image refinement network. The coarse image neural network generates a coarse prediction of imagery for filling the holes of the holey digital image. The image refinement network receives the coarse prediction as input, refines the coarse prediction, and outputs a filled digital image having refined imagery that fills these holes. The image refinement network generates refined imagery using a patch matching technique, which includes leveraging information corresponding to patches of known pixels for filtering patches generated based on the coarse prediction. Based on this, the image completer outputs the filled digital image with the refined imagery.
    Type: Grant
    Filed: May 15, 2018
    Date of Patent: August 25, 2020
    Assignee: Adobe Inc.
    Inventors: Zhe Lin, Xin Lu, Xiaohui Shen, Jimei Yang, Jiahui Yu