3D Model Fusion Using Aerial-Ground Registration and Texture Filling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing reality-based three-dimensional models face issues such as incomplete ground images due to obstacle blockage, geometric distortion, and low accuracy, especially in aerial models, and inefficiencies in ground models that prevent comprehensive city modeling.
Innovation Solution
A method involving feature extraction, registration, triangular patch reconstruction, and texture filling of aerial and ground three-dimensional models to fuse and refine the transition areas, using algorithms like Alpha Shapes and Markov random fields for improved image quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If aerial three-dimensional models are generated using UAV-borne sensors, then the top information of the ground and objects can be captured, but the shot ground image becomes incomplete due to obstacle blockage, resulting in defects such as texture drawing, blurring, broken holes, and geometric model distortion
Solution Approach 1:
The patent segments the three-dimensional modeling process into two distinct parts: aerial three-dimensional models (capturing top information) and ground three-dimensional models (capturing ground texture). By dividing the modeling task, the system can compensate for the incompleteness of aerial images using ground-based models, thereby resolving the issue of obstacle blockage while maintaining complete ground image coverage.
Solution Approach 2:
The patent merges aerial three-dimensional models and ground three-dimensional models into a unified fused three-dimensional model. This combination allows the system to leverage the strengths of both approaches: aerial models provide top-down coverage while ground models fill in areas blocked by obstacles, eliminating defects such as texture drawing, blurring, and broken holes through complementary data integration.
2Measurement precision
If ground three-dimensional models are built using ground close-range photography, then the ground texture can be captured, but the top information of the ground and objects cannot be taken into consideration simultaneously
Solution Approach 1:
The patent combines ground three-dimensional models (which capture ground texture) with aerial three-dimensional models (which capture top information). This merging enables the system to simultaneously consider both ground-level details and overhead views, achieving comprehensive scenario reconstruction that neither approach could accomplish alone.
3Measurement precision
If large-scale city modeling is performed using ground three-dimensional models, then the ground texture can be captured, but the data acquisition time is long and efficiency is low, preventing complete reconstruction of the scenario
Solution Approach 1:
The patent merges aerial and ground three-dimensional modeling approaches to create a hybrid system. The aerial component provides rapid overhead coverage for large-scale areas, while the ground component adds detailed texture information. This combination maintains high modeling accuracy while significantly reducing data acquisition time compared to using ground-based methods alone, enabling complete reconstruction of large-scale scenarios.
Solution Approach 2:
The patent introduces an aerial dimension to the traditional ground-based three-dimensional modeling process. By adding this vertical perspective, the system can efficiently capture large-scale city areas from above while selectively using ground-based photography only where needed, thereby improving overall productivity without sacrificing modeling accuracy.
Data Source
AI summary
Embodiments of the present application disclose a three-dimensional model generating method and apparatus, and an electronic device. The three-dimensional model generating method includes: performing feature extraction on an aerial three-dimensional model and a ground three-dimensional model of a to-be-modeled object respectively to obtain at least three sets of feature vectors; performing registration on the aerial three-dimensional model and the ground three-dimensional model based on the feature vectors, and determining a first transition area of a fused three-dimensional model; reconstructing a triangular patch of the first transition area to obtain a second transition area; and performing texture filling on the second transition area, and determining a target three-dimensional model of the to-be-modeled object.