Patents Examined by Young Lee
  • Patent number: 10992962
    Abstract: A coding distortion removing method is provided for removing coding distortion in two adjacent blocks located on both sides of a block boundary between the two adjacent blocks in a reconstructed image. The method includes deriving a difference of pixel values between a pixel in a first block of the reconstructed image and a pixel in a second block of the reconstructed image adjacent to the first block, deriving an average value of the first quantization parameter and the second quantization parameter, and setting a threshold value to the average value of the first quantization parameter and the second quantization parameter. The method also includes comparing the difference of pixel values with the threshold value, and removing a coding distortion in an area disposed on both sides of the block boundary between the first block and the second block, by applying a filter for coding distortion removal.
    Type: Grant
    Filed: March 19, 2015
    Date of Patent: April 27, 2021
    Assignee: GODO KAISHA IP BRIDGE 1
    Inventors: Zhong Xue, Sheng Mei Shen, Teck Wee Foo, Chak Joo Lee, Shinya Kadono
  • Patent number: 10965954
    Abstract: A picture coding and decoding method includes a picture coding method for coding an input picture and a picture decoding method for decoding coded picture data. The picture coding method includes dividing the input picture into blocks, coding the input picture on a block basis, decoding the coded data of each block on a block basis, removing coding distortion in an area disposed on both sides of a block boundary, each block being adaptively decoded either as a field structure block or a frame structure block, and storing the reconstructed picture as a reference picture. The picture decoding method includes decoding the coded picture data to obtain a reconstructed picture, removing coding distortion in an area disposed on both sides of a block boundary, each block being adaptively decoded either as a field structure block or a frame structure block, and storing the reconstructed picture as a reference picture.
    Type: Grant
    Filed: November 5, 2019
    Date of Patent: March 30, 2021
    Assignee: GODO KAISHA IP BRIDGE 1
    Inventors: Shinya Kadono, Sheng Mei Shen, Teck Wee Foo, Chak Joo Lee, Zhong Xue
  • Patent number: 10944988
    Abstract: A method and apparatus of processing multimedia data that includes a first section received in error are disclosed. Error distribution information corresponding to the first section may be obtained. One of a plurality of error recovery schemes may be selected based on the error distribution information for the first section. The selection may be based at least on whether a channel switch is detected. The first section may be processed based on the selected error recovery scheme.
    Type: Grant
    Filed: July 26, 2017
    Date of Patent: March 9, 2021
    Assignee: Qualcomm Incorporated
    Inventors: Fang Shi, Seyfullah Halit Oguz, Sumeet Singh Sethi, Vijayalakshmi R. Raveendran
  • Patent number: 10939134
    Abstract: A picture coding method provides coding an input picture. The method includes dividing the input picture into a plurality of blocks, coding the input picture on a block basis to obtain coded data of each block, and decoding the coded data of each block on a block basis to obtain a reconstructed picture having a plurality of blocks. The method also includes removing coding distortion in an area disposed on both sides of a block boundary between a first block and an adjacent second block in the reconstructed picture having the plurality of blocks. Each block is adaptively decoded either as a field structure block, including only even field pixels or only odd field pixels, or a frame structure block, including odd field pixels and even field pixels. The method further includes storing the reconstructed picture, for which coding distortion is removed, as a reference picture.
    Type: Grant
    Filed: November 5, 2019
    Date of Patent: March 2, 2021
    Assignee: GODO KAISHA IP BRIDGE 1
    Inventors: Shinya Kadono, Sheng Mei Shen, Teck Wee Foo, Chak Joo Lee, Zhong Xue
  • Patent number: 10834383
    Abstract: When video quality is assessed, a proportion of a damaged data caused by packet loss in a video frame in which the damaged data is located is considered, and other factors are also considered, such as a position of the damaged data in the video frame and impact of a scene in which the video frame is located on the video frame or a video frame located after the video frame. Therefore, accuracy of quality assessment is higher.
    Type: Grant
    Filed: February 22, 2019
    Date of Patent: November 10, 2020
    Assignee: HUAWEI TECHNOLOGIES CO., LTD.
    Inventors: Jie Xiong, Youqing Yang, Yihong Huang
  • Patent number: 10547851
    Abstract: Disclosed are methods for coding and decoding at picture layer and slice layer, and corresponding coder and decoder. The method comprises: coder selecting, according to network transmission state and video characteristic, different constrained loop filter (CLF) and/or constrained intra-prediction (CIP) control modes for one of the following three objects: 1: different video sequences; 2: pictures in different periods in the same video sequence; 3: different slices in the same picture; according to the selected CLF and/or CIP control mode, coding CLF and/or CIP control flag bits in picture layer parameter set and slice layer parameter set; and a decoder parsing slice layer bitstream containing header information, picture layer parameter set and/or slice layer parameter set, and according to the obtained CLF and/or CIP control flag bits, decoding picture and slice by using the corresponding CLF and/or CIP control mode.
    Type: Grant
    Filed: November 1, 2012
    Date of Patent: January 28, 2020
    Assignee: ZTE Corporation
    Inventors: Ming Li, Ping Wu
  • Patent number: 10542261
    Abstract: Techniques and systems are provided for decoding video data. For example, a method of decoding video data includes receiving, by a decoding device, an encoded video bitstream generated according to a first coding protocol, and determining that a base layer is to be provided to the decoding device as part of an external bitstream that is different from the received encoded video bitstream. The external bitstream is provided to the decoding device by an external source and is not provided as part of the encoded video bitstream. The base layer is encoded using a second coding protocol that is different from the first coding protocol. The method further includes determining to ignore a profile, tier, and level syntax structure assigned to the base layer in the video parameter set in response to determining the base layer is to be provided to the decoding device as part of an external bitstream.
    Type: Grant
    Filed: January 30, 2018
    Date of Patent: January 21, 2020
    Assignee: QUALCOMM Incorporated
    Inventors: Ye-Kui Wang, Fnu Hendry, Adarsh Krishnan Ramasubramonian
  • Patent number: 10425576
    Abstract: A drone including a flying unit configured to generate a lift force for flying; a communication unit configured to receive sensing data obtained in a mobile terminal from the mobile terminal; a camera configured to capture a specific object; and a controller configured to recognize at least one of a movement change of the specific object and a status change of the specific object based on the received sensing data, change a capturing composition of the specific object based on the recognized at least one of the movement change and the status change, and capture the specific object via the camera based on the changed capturing composition.
    Type: Grant
    Filed: December 21, 2016
    Date of Patent: September 24, 2019
    Assignee: LG ELECTRONICS INC.
    Inventors: Minjae Kim, Hokyun Kang
  • Patent number: 10404990
    Abstract: An apparatus for coding video information according to certain aspects includes a memory and a processor configured to: determine a value of a first flag indicative of whether preceding pictures that precede a current picture in a current layer in decoding order are used for temporal motion vector prediction (TMVP) of subsequent pictures that follow the current picture in the current layer in decoding order, the first flag associated with the current picture; and refrain from using the preceding pictures in the current layer as reference pictures for TMVP of the current picture, in response to determining that the value of the first flag indicates that the preceding pictures in the current layer should not be used for TMVP of the subsequent pictures in the current layer.
    Type: Grant
    Filed: January 30, 2018
    Date of Patent: September 3, 2019
    Assignee: Qualcomm Incorporated
    Inventors: Fnu Hendry, Adarsh Krishnan Ramasubramonian, Ye-Kui Wang
  • Patent number: 10297008
    Abstract: Methods, systems, and computer programs for improving compressed image chroma information. In one aspect of the invention, a resolution for a red color component of a color video image is used that is higher than the resolution for a blue color component of the color video image. Another aspect includes utilizing a lower or higher value of a quantization parameter (QP) for one or more chroma channels as compared to the luminance channel. Another aspect is use of a logarithmic representation of a video image to benefit image coding. Another aspect uses more than two chroma channels to represent a video image.
    Type: Grant
    Filed: December 14, 2011
    Date of Patent: May 21, 2019
    Assignee: Dolby Laboratories Licensing Corporation
    Inventor: Gary A. Demos
  • Patent number: 10225577
    Abstract: The present invention presents fully noncausal predictive encoding and decoding methods for image, video and other signal coding. The presented noncausal predictive image coding methods largely reduce the prohibitive computational cost of the prior invention. The presented noncausal signal encoding method comprises: (1) splitting the source signal into a plurality of noncausal coding units; (2) extending each noncausal coding unit with the selected extension type; and (3) encoding each noncausal coding unit with the selected intra-unit noncausal predictor and intra-unit noncausal predictive encoding method.
    Type: Grant
    Filed: July 23, 2015
    Date of Patent: March 5, 2019
    Inventor: Shidong Chen
  • Patent number: 10178386
    Abstract: An image decoding apparatus that includes a motion compensation prediction circuit configured to conduct motion compensation prediction for each of blocks to be decoded by using the reconstructed image, an inverse transformation circuit configured to conduct inverse orthogonal transformation for the data of the blocks to be decoded, and a determination circuit configured to determine a filtering strength and whether or not to conduct filtering, with respect to each of the boundaries. In addition, the determining circuit is configured to determine filtering is conducted when at least one of the two adjacent blocks is intra-coded, and filtering is not conducted when both of the two adjacent blocks are not intra-coded, a non-zero transformation coefficient is not coded in both of the two adjacent blocks, the two adjacent blocks are predicted by the same reference frame, and an absolute value of a difference between motion vectors of the two adjacent blocks is smaller than a specified threshold value.
    Type: Grant
    Filed: November 9, 2017
    Date of Patent: January 8, 2019
    Assignee: Dolby Laboratories Licensing Corporation
    Inventors: Shijun Sun, Shawmin Lei, Hiroyuki Katata
  • Patent number: 10158865
    Abstract: Methods, systems, and computer programs for improving compressed image chroma information. In one aspect of the invention, a resolution for a red color component of a color video image is used that is higher than the resolution for a blue color component of the color video image. Another aspect includes utilizing a lower or higher value of a quantization parameter (QP) for one or more chroma channels as compared to the luminance channel Another aspect is use of a logarithmic representation of a video image to benefit image coding. Another aspect uses more than two chroma channels to represent a video image.
    Type: Grant
    Filed: August 17, 2015
    Date of Patent: December 18, 2018
    Assignee: Dolby Laboratories Licensing Corporation
    Inventor: Gary A. Demos
  • Patent number: 10110908
    Abstract: Methods, systems, and computer programs for improving compressed image chroma information. In one aspect of the invention, a resolution for a red color component of a color video image is used that is higher than the resolution for a blue color component of the color video image. Another aspect includes utilizing a lower or higher value of a quantization parameter (QP) for one or more chroma channels as compared to the luminance channel. Another aspect is use of a logarithmic representation of a video image to benefit image coding. Another aspect uses more than two chroma channels to represent a video image.
    Type: Grant
    Filed: August 17, 2015
    Date of Patent: October 23, 2018
    Assignee: Dolby Laboratories Licensing Corporation
    Inventor: Gary A. Demos
  • Patent number: 10068140
    Abstract: A vehicle movement parameter, such as ego-speed, is estimated using real-time images captured by a single camera. The captured images may be analyzed by a pre-trained convolutional neural network to estimate vehicle movement based on monocular video data. The convolutional neural network may be pre-trained using filters from a synchrony autoencoder that were trained using unlabeled video data captured by the vehicle's camera while the vehicle was in motion. A parameter corresponding to the estimated vehicle movement may be output to the driver or to a driver assistance system for use in controlling the vehicle.
    Type: Grant
    Filed: December 2, 2016
    Date of Patent: September 4, 2018
    Assignees: Bayerische Motoren Werke Aktiengesellschaft, NAUTO, Inc.
    Inventors: Ludmila Levkova, Koyel Banerjee
  • Patent number: 10051269
    Abstract: Systems and methods may be provided for determining whether or not to output the “decoded pictures yet to be output” (DPYTBO) after decoding a random access point (RAP) picture of the second of two consecutive coded video sequences in a bitstream. The DPYTBO pictures may reside in a portion of memory coupled to a video decoder. The portion of memory may correspond to the Decoder Picture Buffer (DPB) of the video decoder. The picture resolution in the bitstream is allowed to change only at the start of a CVS. The first coded picture in a coded video sequence (CVS) may be a RAP picture. The second CVS may or may not have a picture resolution that is different from the picture resolution corresponding to the first CVS.
    Type: Grant
    Filed: August 24, 2017
    Date of Patent: August 14, 2018
    Assignee: Cisco Technology, Inc.
    Inventors: Arturo A. Rodriguez, Anil Kumar Katti, Hsiang-Yeh Hwang
  • Patent number: 10015524
    Abstract: A frame coding unit generates individual picture signal information from a moving picture signal, and outputs a frame code value which is a numeric value to be obtained as a result of coding picture signal information of each frame. A variable length coding unit decomposes the frame code value into unit frame code values which are basic units of coding, converts the unit frame code values into frame code words using only a single code table, and constructs a frame stream by combining the converted frame code words. A multiplexing unit multiplexes a header stream which is constructed through a method equivalent to a conventional method and the frame stream, and constructs a picture coded signal.
    Type: Grant
    Filed: September 29, 2017
    Date of Patent: July 3, 2018
    Assignee: PANASONIC INTELLECTUAL PROPERTY CORPORATION OF AMERICA
    Inventors: Shinya Kadono, Makoto Hagai
  • Patent number: 10015517
    Abstract: A coding distortion removal method is provided for removing coding distortion in an area disposed on both sides of a block boundary between a first block and an adjacent second block in a picture having a plurality of blocks, each block being adaptively coded either as a field structure block, comprising only even field pixels or comprising only odd field pixels, or a frame structure block, comprising odd field pixels and even field pixels. The method includes detecting whether the block boundary between the first block and the adjacent second block is a block boundary between a field structure block and a frame structure block, and performing a coding distortion removal process on the first block and the adjacent second block. The coding distortion removal process includes determining whether coding distortion removal is needed, determining the number of pixels to be processed, and removing coding distortion.
    Type: Grant
    Filed: November 14, 2017
    Date of Patent: July 3, 2018
    Assignee: GODO KAISHA IP BRIDGE 1
    Inventors: Shinya Kadono, Sheng Shen, Teck Wee Foo, Chak Joo Lee, Zhong Xue
  • Patent number: 10008003
    Abstract: A process generates a lookup table to estimate spatial depth in a visual scene. The process identifies subsets of illuminators of a camera system with image sensors and illuminators. The image sensors are associated with multiple pixels. For each pixel, and for each of multiple depths from the pixel, the process simulates a virtual surface at the depth. For each subset of the subsets of illuminators, the process simulates illumination of the virtual surface from the subset and determines an expected light intensity at the pixel from light reflected from the virtual surface due to the simulated illumination. The process forms intensity information from the expected light intensity determined for the pixel for each of the depths and each of the subsets. The process constructs a lookup table comprising the intensity information. The lookup table associates the intensity information for each pixel with the respective depth and the respective subset.
    Type: Grant
    Filed: April 30, 2017
    Date of Patent: June 26, 2018
    Assignee: GOOGLE LLC
    Inventors: Ghulam Ahmed Kirmani, Andrea Colaco, Dongeek Shin
  • Patent number: 9986265
    Abstract: A frame coding unit generates individual picture signal information from a moving picture signal, and outputs a frame code value which is a numeric value to be obtained as a result of coding picture signal information of each frame. A variable length coding unit decomposes the frame code value into unit frame code values which are basic units of coding, converts the unit frame code values into frame code words using only a single code table, and constructs a frame stream by combining the converted frame code words. A multiplexing unit multiplexes a header stream which is constructed through a method equivalent to a conventional method the frame stream, and constructs a picture coded signal.
    Type: Grant
    Filed: September 29, 2017
    Date of Patent: May 29, 2018
    Assignee: PANASONIC INTELLECTUAL PROPERTY CORPORATION OF AMERICA
    Inventors: Shinya Kadono, Makoto Hagai