Image processing method, computer device, and storage medium
By reconstructing a 3D scene model and determining the viewpoint rendering trajectory, images or videos from different perspectives are generated, solving the problem that traditional cameras can only shoot from a single perspective and improving the user experience.
WO2026143405A1PCT designated stage Publication Date: 2026-07-09ARASHI VISION INC
Patent Information
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- ARASHI VISION INC
- Filing Date
- 2024-12-31
- Publication Date
- 2026-07-09
AI Technical Summary
Technical Problem
Traditional cameras can only capture images from a single perspective, which cannot meet users' needs for a full-view shot of highlights.
Method used
By acquiring the image to be processed, a 3D scene model is reconstructed, the viewpoint rendering trajectory is determined, and the target image or video is generated to achieve image display from different perspectives.
Benefits of technology
It enables multi-view display based on a single image, improving the user experience and ensuring geometric consistency of image content from different perspectives.
✦ Generated by Eureka AI based on patent content.
Smart Images
Figure CN2024144111_09072026_PF_FP_ABST
Abstract
Disclosed in the embodiments of the present application are an image processing method, a computer device, and a storage medium. The method comprises: acquiring a three-dimensional scene model reconstructed on the basis of an image to be processed; on the basis of the three-dimensional scene model, determining a viewing angle rendering trajectory; and on the basis of the viewing angle rendering trajectory, generating a corresponding target style image or video. In the method, a new viewing angle is determined by means of a three-dimensional scene model reconstructed on the basis of an image to be processed. Compared with an original photographing viewing angle of a camera or a photographing viewing angle obtained by means of moving the camera, the three-dimensional scene model is more stereoscopic and comprises a panoramic viewing angle image of the image to be processed; thus, the new viewing angle can be flexibly selected and determined, such that the new viewing angle may have a larger variation range; furthermore, different viewing angle pictures may be obtained on the basis of one image, thereby improving the user experience.
Need to check novelty before this filing date? Find Prior Art