Modeling method and electronic equipment

By obtaining standard images of joint objects at multiple viewing angles and states, generating a deformed Gaussian field and performing component segmentation processing, the limitations of movable component objects modeling and motion analysis in real scenes in the prior art are solved, and high-precision multi-state and multi-component modeling is achieved.

CN120599148AActive Publication Date: 2025-09-05BEIJING HUMANOID ROBOTICS INNOVATION CENTER CO LTD
3 Cites 1 Cited by

Patent Information

Application Number
CN202510771162.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-06-10
Publication Date
2025-09-05
Estimated Expiration
2045-06-10

AI Technical Summary

Technical Problem

The prior art has limitations when performing three-dimensional modeling and motion analysis of objects with movable components in real scenarios, especially based on the explicit geometric representation scheme, the implicit shape representation scheme cannot accurately decompose the independent motion of multiple components, and the combination of NeRF and 3DGS schemes cannot cope with multi-moving components and multi-interaction states.

Method used

By obtaining standard images of joint objects in multiple viewing angles and multiple states, a deformed Gaussian field is generated, component segmentation processing is performed, the motion parameters of the moving parts are determined, and modeled, and a target Gaussian field is generated to characterize the geometric structure, appearance and motion characteristics of the moving parts.

Benefits of technology

It realizes the modeling accuracy and efficiency of joint objects in multiple states without manual labeling or pre-defined number of moving parts, and generates a multi-state and multi-component target Gaussian field with clear moving parts and clear boundaries.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention provides a modeling method and electronic equipment. The method comprises the following steps: acquiring standard images of a joint object at a plurality of visual angles and a plurality of states; according to the standard image, a deformation Gaussian field of the joint object in each state is generated, and the deformation Gaussian field is used for representing the geometric structure, appearance and motion characteristics of the joint object in the corresponding state; according to the deformation Gaussian field, part segmentation processing is carried out, and a target Gaussian field of each moving part in the joint object in each state is generated; according to the target Gaussian fields of all the moving parts in the initial state and all the target states, determining the moving parameters of all the moving parts in all the target states, and according to the moving parameters of all the moving parts in all the target states, carrying out modeling on all the moving parts, and the target model of the joint object in each target state is obtained according to the modeling result of each moving part, so that the modeling precision and efficiency of the joint object in various states are improved.
Need to check novelty before this filing date? Find Prior Art