公开(公告)号：US12165293B2

公开(公告)日：2024-12-10

申请号：US17487538

申请日：2021-09-28

Applicant: SAMSUNG ELECTRONICS CO., LTD.

Inventor： Kaimo Lin ,  Hamid Sheikh

IPC: G06T5/77 ,  G06T3/4053 ,  G06T5/50 ,  G06T7/33 ,  G06V10/54

Abstract: An apparatus for processing image data may include: a memory storing instructions; and a processor configured to execute the instructions to: extract a target image patch including a target object, from a captured image; obtain a plurality of landmark features from the target image patch; align the plurality of landmark features of the target image patch with a plurality of reference landmark features in a template image patch including the same target object; and when the plurality of landmark features are aligned with the plurality of reference landmark features, transfer texture details of the target object in the template image patch to the target object in the target image patch.

公开(公告)号：US20230245330A1

公开(公告)日：2023-08-03

申请号：US17589014

申请日：2022-01-31

Applicant: Samsung Electronics Co., Ltd.

Inventor： Zeeshan Nadir ,  Numair Khan ,  Hamid Sheikh

IPC: G06T7/33 ,  G06T7/50 ,  G06T5/00 ,  G06T5/40 ,  G06T3/00

CPC classification number: G06T7/337 ,  G06T7/50 ,  G06T5/005 ,  G06T5/40 ,  G06T3/0075 ,  G06T2207/20084 ,  G06T2207/20081 ,  G06T2207/30201

Abstract: A method includes aligning landmark points between multiple distorted images to generate multiple aligned images, where the multiple distorted images exhibit perspective distortion in at least one face appearing in the multiple distorted images. The method also includes predicting a depth map using a disparity estimation neural network that receives the multiple aligned images as input. The method further includes generating a warp field using a selected one of the multiple aligned images. The method also includes performing a two-dimensional (2D) image projection on the selected aligned image using the depth map and the warp field to generate an undistorted image. In addition, the method includes filling in one or more missing pixels in the undistorted image using an inpainting neural network to generate a final undistorted image.

公开(公告)号：US20220020126A1

公开(公告)日：2022-01-20

申请号：US17103850

申请日：2020-11-24

Applicant: Samsung Electronics Co., Ltd.

Inventor： Ruiwen Zhen ,  John Glotzbach ,  Hamid Sheikh

IPC: G06T5/00 ,  G06T5/50 ,  G06T5/20

Abstract: A method includes obtaining multiple images of a scene including a first image associated with a first exposure value and a second image associated with a second exposure value. Each image includes image data in each of multiple channels of a color space, and the first exposure value is greater than the second exposure value. The method also includes decomposing each channel of each image into a base layer and a detail layer. The method further includes generating for each image, a blending weight map for the base layer and a blending weight map for the detail layer. The method includes combining each base layer and each detail layer of each channel of each image based on the blending weight map for the base layer for the image and the blending weight map for the detail layer of the image to obtain an HDR image of the scene.

公开(公告)号：US11430094B2

公开(公告)日：2022-08-30

申请号：US17103850

申请日：2020-11-24

Applicant: Samsung Electronics Co., Ltd.

Inventor： Ruiwen Zhen ,  John Glotzbach ,  Hamid Sheikh

IPC: G06T5/00 ,  G06T5/50 ,  G06T5/20

Abstract: A method includes obtaining multiple images of a scene including a first image associated with a first exposure value and a second image associated with a second exposure value. Each image includes image data in each of multiple channels of a color space, and the first exposure value is greater than the second exposure value. The method also includes decomposing each channel of each image into a base layer and a detail layer. The method further includes generating for each image, a blending weight map for the base layer and a blending weight map for the detail layer. The method includes combining each base layer and each detail layer of each channel of each image based on the blending weight map for the base layer for the image and the blending weight map for the detail layer of the image to obtain an HDR image of the scene.