公开(公告)号：US20240177379A1

公开(公告)日：2024-05-30

申请号：US18551783

申请日：2022-03-21

Applicant: BEIJING ZITIAO NETWORK TECHNOLOGY CO., LTD.

Inventor： Lijuan YU ,  Qili DENG

IPC: G06T11/20 ,  G06T7/13 ,  G06T11/60

CPC classification number: G06T11/203 ,  G06T7/13 ,  G06T11/60 ,  G06T2207/30201 ,  G06T2210/44

Abstract: The present disclosure provides an image processing method and apparatus, a device and a storage medium. The method comprises: first, determining a first contour point and a second contour point of a target object on an image to be processed, wherein the first contour point belongs to points on a first contour line of the target object, and the second contour point belongs to points on a second contour line of the target object; secondly, constructing, on the basis of the first contour point and the second contour point, a grid corresponding to the target object, and shifting the second contour point in the grid on the basis of a deformation parameter corresponding to the target object, to obtain a deformed grid; and then, generating, on the basis of the deformed grid, a deformed image corresponding to the target object. Hence, according to embodiments of the present disclosure, the grid of the target object on an image to be processed is deformed, such that the image processing function is realized, and the display effect of the target object on said image can be improved, thereby improving user experience.

公开(公告)号：US20250131648A1

公开(公告)日：2025-04-24

申请号：US18688212

申请日：2022-08-31

Applicant: Beijing Zitiao Network Technology Co., Ltd.

Inventor： Zhixing CHEN ,  Qili DENG ,  Zhichao LIU

IPC: G06T17/00

Abstract: A three-dimensional human head reconstruction method, an electronic device and a non-transient computer-readable storage medium are provided. The method includes: acquiring a target portrait image; inputting the target portrait image into a target model to obtain an output result of the target model; wherein the target model is obtained by pre-training with a plurality of training samples which are generated according to a sample portrait image and a sample three-dimensional human head model, the sample three-dimensional human head model is obtained by iteratively fitting a standard three-dimensional human face statistical model according to two-dimensional feature information related to a portrait in the sample portrait image, and the two-dimensional feature information includes human face feature points and a human head projection contour line; and generating, according to the output result, a target three-dimensional human head model corresponding to the target portrait image.