3D Building Model Generation from Textured Mesh Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for generating 3D building models for location-based services are inefficient, prone to errors due to misalignment of different data sources, and inaccurate in assuming standard building features, leading to challenges in detecting and modeling 3D objects accurately.
Innovation Solution
A method involving processing textured 3D mesh data to generate 2D depth images, filtering out mesh data below a threshold height, and using clustering techniques and model-fitting processes to create accurate partial and complete 3D models of building facades, which are then aligned with their corresponding objects in an application.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If semi-manual processes are used for generating 3D building models, then some level of accuracy can be achieved, but the process is inefficient and not scalable
Solution Approach 1:
The system automatically processes textured 3D mesh data through filtering, depth image generation, and model fitting without manual intervention. The automated workflow includes: filtering mesh data by height thresholds, generating 2D depth images, identifying building facades through image processing, and fitting 3D models automatically, thereby achieving both accuracy and scalability
Solution Approach 2:
The system changes the state of 3D mesh data by converting it into 2D depth images and then into identified building facades. This parameter transformation enables automated processing while maintaining accuracy through systematic application of filtering thresholds and model fitting algorithms
2Loss of information
If multiple data sources (street level and satellite imagery) are used for generating 3D models, then more comprehensive information can be obtained, but misalignment errors occur
Solution Approach 1:
The system segments the processing into distinct stages: first filtering mesh data by height, then generating depth images, followed by facade identification and model fitting. This segmentation allows each stage to focus on specific aspects, reducing cumulative alignment errors that would occur when processing all data sources simultaneously
Solution Approach 2:
The system uses 2D depth images as an intermediary representation between the raw 3D mesh data and the final 3D building models. This intermediate format facilitates accurate alignment and reduces errors by providing a standardized representation that bridges different data sources
3Device complexity
If standard building assumptions (vertical walls, flat roofs, flat ground) are made, then processing is simplified, but accuracy decreases for non-standard buildings
Solution Approach 1:
The system dynamically adapts to different building types by using data-driven threshold selection and adaptive model fitting. Instead of assuming fixed building characteristics, the system determines thresholds based on the actual mesh data distribution and fits models that conform to the observed geometry, whether standard or nonstandard
Solution Approach 2:
The system applies different processing parameters and thresholds to different regions of the 3D mesh data based on local characteristics. By analyzing the specific geometry and texture of each building area, the system tailors the filtering and model fitting parameters to match local building features rather than applying uniform assumptions
4Loss of information
If all mesh data is processed without filtering, then complete information is retained, but noise and irrelevant data reduce model accuracy
Solution Approach 1:
The system extracts only the relevant portions of the 3D mesh data by filtering out elements below a determined height threshold. This extraction process removes noise and irrelevant ground-level mesh data while preserving the building facade information needed for accurate 3D model generation
Data Source
AI summary
An approach is provided for identifying objects present in mesh representation of a geo-location, generating accurate 3D models for the objects, and aligning the 3D models to their corresponding objects in an application. The approach comprises processing and/or facilitating a processing of textured three-dimensional mesh data in one or more regions of interest to cause, at least in part, a generation of at least one two-dimensional depth image representation. The approach further comprises causing, at least in part, a filtering of the textured three-dimensional mesh data in the one or more regions of interest to remove mesh data below at least one threshold height based, at least in part, on the at least one two-dimensional depth image representation. Additionally, the approach comprises processing and/or facilitating a processing of the filtered textured three-dimensional mesh data to cause, at least in part, a generation of at least one partial three-dimensional model, including one or more upper facades above the at least one threshold height, of one or more objects located within the one or more regions of interest.