Depth-Based 3D Modeling Grid Subdivision for Real-Time Rendering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing 3D modeling methods face challenges in rendering high-resolution, free-view 3D models in real-time while maintaining a favorable display effect, particularly due to high computational loads and distortion issues during zoom-in and zoom-out operations.
Innovation Solution
A depth information-based modeling method that receives depth image data and color image data to generate 3D grids and uniform sampling grids, which are further divided into sub-grids based on variance, allowing for the determination of points on the edges of these grids to create triangle meshes, thereby enhancing 3D modeling efficiency and reducing computational load.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If traditional 3D modeling methods are used to generate high-resolution free-view 3D models, then the display effect is favorable, but the computational load is high and real-time rendering is difficult
Solution Approach 1:
The patent divides the 3D model representation into two separate components: depth maps that capture geometric structure and color maps that capture visual appearance. This segmentation allows independent processing of structural and visual information, reducing computational complexity while maintaining high-quality rendering capabilities for free-view 3D models
Solution Approach 2:
The patent uses depth maps as a simplified copy or representation of the 3D model's geometric structure. Instead of processing complex mesh data for all rendering operations, the system uses depth maps as a lightweight proxy that can be efficiently processed in real-time while still enabling high-quality reconstruction and rendering
2Productivity
If MVC method is used for 3D modeling with view synthesis, then real-time rendering is easier, but zoom-in and zoom-out operations cause distortion
Solution Approach 1:
The patent transforms depth information from non-linear representation to linear representation by changing the depth encoding parameters. This parameter transformation enables accurate zoom-in and zoom-out operations without distortion, as linear depth values allow proportional scaling while maintaining geometric accuracy across different view distances
3Ease of operation
If feature points and meshes are used for 3D model construction, then user interaction is easy, but computational load is large for real-time rendering
Solution Approach 1:
The patent extracts only the essential geometric information needed for 3D representation by using depth maps, discarding the complex mesh structures and feature point data. This extraction approach maintains ease of user interaction for model construction while dramatically reducing the computational load for real-time rendering operations
Data Source
AI summary
A depth information-based modeling method, a graphic processing apparatus and a storage medium are provided. The depth information-based modeling method includes receiving a plurality of depth image data, each of which has depth information, and a plurality of color images, each of which color information. In the method, a plurality of 3D grids is obtained according to the color information, depth information and a plurality of uniform sampling grids. Each of the uniform sampling grids is further divided into a plurality of sub-grids. At least one point on each of the 3D grids is determined by the depth information, and the triangle meshes are generated according to the points.