3D Building Geometry Reconstruction from Ground-Level Imagery
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing 3D mapping technologies for urban cities suffer from limited texture resolution, geometry quality, high costs, time consumption, and lack of real-time data analytics, making them inefficient for consumer and commercial use cases.
Innovation Solution
A system and method for generating 3D building models using ground-level and orthogonal imagery, involving image processing and correlation with geo-referenced data to create accurate, textured 3D models.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If aerial imagery or specialized camera-equipped vehicles are used to generate 3D maps, then 3D building models can be created, but the texture resolution and geometry quality are limited
Solution Approach 1:
Instead of using aerial imagery to create 3D models, the patent inverts the approach by using ground-level imagery captured from street view to generate and texture 3D building models. This inversion allows for higher texture resolution as ground-level cameras can capture detailed façade surfaces up close, while the 3D geometry is derived from multiple ground-level perspectives rather than aerial views.
Solution Approach 2:
The patent introduces an intermediary process that correlates ground-level images with orthogonal imagery and uses image processing algorithms to extract geometric features and textures. This intermediary step bridges the gap between simple ground-level photography and high-quality 3D building models, enabling detailed texture mapping without requiring complex aerial mapping systems.
2Manufacturing precision
If aerial imagery or specialized camera-equipped vehicles are used to generate 3D maps, then 3D building models can be created, but the geometry quality is limited
Solution Approach 1:
The patent performs preliminary actions by capturing ground-level imagery and processing it to extract geometric features and building boundaries before generating the 3D models. By pre-processing the ground-level images to identify key geometric elements such as building corners, roof lines, and façade features, the system accelerates the subsequent 3D reconstruction process while maintaining high geometry quality.
Solution Approach 2:
The patent transitions from two-dimensional ground-level images to three-dimensional building models by using multiple ground-level perspectives and correlating them with orthogonal imagery. This dimensional transformation leverages the vertical dimension captured in ground-level photos to reconstruct 3D geometry, achieving high precision without the time consumption of traditional aerial mapping methods.
3Measurement precision
If traditional 3D mapping methods are used, then 3D building models can be generated, but the cost is high
Solution Approach 1:
The patent uses ground-level imagery as a copy or alternative data source instead of expensive aerial photography or specialized mapping equipment. By capturing and processing readily available ground-level images, the system creates accurate 3D building models at a fraction of the cost of traditional methods, while maintaining measurement precision through sophisticated image correlation and processing algorithms.
4Adaptability or versatility
If traditional 3D mapping methods are used, then 3D building models can be generated, but updates are difficult
Solution Approach 1:
The patent implements a dynamic update mechanism where ground-level imagery can be continuously captured and processed to update 3D building models in real-time. As new ground-level images become available, the system can automatically correlate them with existing models and update geometries and textures, making the mapping system adaptable and easy to maintain without increasing overall system complexity.
5Productivity
If traditional 3D mapping methods are used, then 3D building models can be generated, but real-time image data analytics are not provided
Solution Approach 1:
The patent creates a multi-functional system that not only generates 3D building models but also provides real-time image data analytics. The same ground-level imagery processing pipeline that creates 3D models also extracts and analyzes image data such as building attributes, environmental features, and urban characteristics, enabling the system to serve multiple purposes without proportionally increasing complexity.
Data Source
AI summary
Systems and methods are described for creating three dimensional models of building objects by creating a point cloud from a plurality of input images, defining edges of the building object's surfaces represented by the point cloud, creating simplified geometries of the building object's surfaces and constructing a building model based on the simplified geometries. Input images may include ground, orthographic, or oblique images. The resultant model may be scaled according to correlation with select image types and textured.