Patents by Inventor Benshun Yi

Benshun Yi has filed for patents to protect the following inventions. This listing includes patent applications that are pending as well as patents that have already been granted by the United States Patent and Trademark Office (USPTO).

  • Patent number: 9196024
    Abstract: Embodiments of the present invention disclose a method and an apparatus for enhancing color, and relate to the image processing field, so as to solve a case of over enhancement on a region having high luminance in an original image which is caused after enhancement processing is performed. A specific solution is: obtaining a luminance component of a currently processed image; performing Gaussian filtering processing on the luminance component to obtain an illumination component; obtaining a first reflection component of the currently processed image according to the luminance component and the illumination component; performing enhancement adjustment on the first reflection component to obtain a second reflection component; obtaining a luminance gain according to the luminance component and the second reflection component; and performing enhancement processing on the currently processed image, to obtain a first enhanced image. The present invention is applicable to a process of color enhancement.
    Type: Grant
    Filed: July 15, 2014
    Date of Patent: November 24, 2015
    Assignee: Huawei Technologies Co., Ltd.
    Inventors: Diwang Weng, Jinsheng Xiao, Benshun Yi
  • Patent number: 8913185
    Abstract: The present invention provides a multi-level bidirectional motion estimation method and device searching for matched blocks in a current frame and a previous frame respectively by using M1×M1 blocks of a to-be-interpolated frame as a mirror center, where the first to-be-determined matched block pair is a pair of matched blocks with a minimum sum of binary absolute difference SHAD value among all pairs of matched blocks searched out in the current frame and the previous frame; and if determined that the SHAD value of the first to-be-determined matched block pair is greater than or equal to an SHAD threshold, splitting the M1×M1 blocks into M2×M2 blocks, and searching for matched blocks again in the current frame and the previous frame respectively by using the M2×M2 blocks of the to-be-interpolated frame as a mirror center. Therefore, accuracy of motion estimation can be improved effectively, and image quality can be improved.
    Type: Grant
    Filed: July 15, 2014
    Date of Patent: December 16, 2014
    Assignee: Huawei Technologies Co., Ltd.
    Inventors: Xiaojun Wu, Benshun Yi, Jinsheng Xiao
  • Publication number: 20140327819
    Abstract: The present invention provides a multi-level bidirectional motion estimation method and device. searching for matched blocks in a current frame and a previous frame respectively by using M1×M1 blocks of a to-be-interpolated frame as a mirror center, where the first to-be-determined matched block pair is a pair of matched blocks with a minimum sum of binary absolute difference SHAD value among all pairs of matched blocks searched out in the current frame and the previous frame; and if determined that the SHAD value of the first to-be-determined matched block pair is greater than or equal to an SHAD threshold, splitting the M1×M1 blocks into M2×M2 blocks, and searching for matched blocks again in the current frame and the previous frame respectively by using the M2×M2 blocks of the to-be-interpolated frame as a mirror center. Therefore, accuracy of motion estimation can be improved effectively, and image quality can be improved.
    Type: Application
    Filed: July 15, 2014
    Publication date: November 6, 2014
    Inventors: Xiaojun Wu, Benshun Yi, Jinsheng Xiao
  • Publication number: 20140328539
    Abstract: Embodiments of the present invention disclose a method and an apparatus for enhancing color, and relate to the image processing field, so as to solve a case of over enhancement on a region having high luminance in an original image which is caused after enhancement processing is performed. A specific solution is: obtaining a luminance component of a currently processed image; performing Gaussian filtering processing on the luminance component to obtain an illumination component; obtaining a first reflection component of the currently processed image according to the luminance component and the illumination component; performing enhancement adjustment on the first reflection component to obtain a second reflection component; obtaining a luminance gain according to the luminance component and the second reflection component; and performing enhancement processing on the currently processed image, to obtain a first enhanced image. The present invention is applicable to a process of color enhancement.
    Type: Application
    Filed: July 15, 2014
    Publication date: November 6, 2014
    Inventors: Diwang Weng, Jinsheng Xiao, Benshun Yi