3D Model Geometric Unit Segmentation for Lower Rendering Costs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Three-dimensional stereoscopic models require a large number of parameters, leading to high rendering and storage costs due to the need for numerous surface patches or point clouds.
Innovation Solution
Convert the model into a representation based on three-dimensional geometric units such as cylinders, spheres, and cubes, using machine learning models to segment and determine the geometric units, reducing the number of required parameters by assembling these units to form a new model.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a three-dimensional stereoscopic model is represented using numerous surface patches or point clouds, then model accuracy is improved, but the number of parameters increases leading to high rendering and storage costs
Solution Approach 1:
The patent segments the complex three-dimensional model into multiple simple geometric units (cubes, spheres, cylinders, cones). Each geometric unit is defined by a limited set of parameters (position, size, shape type), replacing the need for numerous surface patches or point clouds. This segmentation maintains model accuracy while dramatically reducing the total parameter count.
Solution Approach 2:
The patent uses simple geometric primitives (standardized cube, sphere, cylinder, cone shapes) as building blocks to copy and reconstruct complex model features. These standardized geometric units serve as templates that can be repeatedly instantiated with different parameters to represent various parts of the three-dimensional model, reducing complexity while preserving accuracy.
2Quantity of substance
If the number of parameters in a three-dimensional model is reduced by using geometric units, then rendering and storage costs decrease, but model complexity increases due to the need for segmentation and assembly
Solution Approach 1:
The patent automatically segments the input three-dimensional model into distinct geometric units using image processing and machine learning algorithms. This automated segmentation handles the complexity of decomposition, allowing the system to break down complex models into simple geometric primitives without requiring manual intervention, thus managing construction complexity while reducing parameter count.
Solution Approach 2:
The patent transforms the representation parameters of the model from detailed surface-level descriptions (millions of vertices and faces) to simplified geometric parameters (position coordinates, dimensions, shape types). This parameter transformation reduces the total number of parameters while the system manages the complexity of converting between representations through automated algorithms.
Data Source
AI summary
The present disclosure relates to a stereoscopic model processing method, apparatus, device and medium. The method includes: acquiring a first stereoscopic model; segmenting the first stereoscopic model based on a plurality of types of preset three-dimensional geometric units, obtaining a type of a three-dimensional geometric unit to which a point of the first stereoscopic model belongs, and obtaining each target component corresponding to the first stereoscopic model; determining a target three-dimensional geometric unit corresponding to the each target component according to point information of the each target component, wherein the point information of the each target component includes spatial coordinates of a point of the each target component and a type of a three-dimensional geometric unit to which the point of the each target component belongs; and determining a second stereoscopic model based on the target three-dimensional geometric unit corresponding to the each target component.