Automated 3D Modeling with Real Façade Texture Occlusion Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current three-dimensional modeling techniques face challenges in accurately representing building façades without obstructions, as oblique images often contain occlusions and require manual manipulation, which is time-consuming and costly.
Innovation Solution
A method for automatically generating three-dimensional models using geo-referenced oblique images, where the geographical position of façade textures is determined using wire-frame data, and occlusions are corrected by analyzing and applying unoccluded image characteristics to form a mosaic image, ensuring accurate representation of building façades.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If manual manipulation of oblique images is used to form unobstructed façade views, then accuracy of building representation is improved, but time consumption and operational complexity increase significantly
Solution Approach 1:
The system automatically selects and processes oblique images without requiring manual manipulation. The computer automatically identifies suitable images, corrects occlusions, and applies textures to the three-dimensional model, eliminating the need for expert manual intervention while maintaining high accuracy in building representation.
Solution Approach 2:
The patent replaces manual mechanical manipulation of images with automated computer-based image processing algorithms. The system uses automated occlusion detection, image selection, and texture application processes that substitute human operators with computational methods, significantly reducing time consumption while preserving accuracy.
2Productivity
If automated image selection is used to reduce manual intervention, then productivity increases, but accuracy of façade texture representation may deteriorate
Solution Approach 1:
The system incorporates automated feedback mechanisms where the computer analyzes image characteristics, detects occlusions, and selects the most suitable images based on predefined criteria. This automated feedback loop ensures that only high-quality, unobstructed images are selected and applied, maintaining texture accuracy while enabling rapid automated processing of multiple buildings.
Solution Approach 2:
The patent changes key parameters of the image processing workflow by implementing automated threshold-based occlusion detection, automated image quality assessment, and algorithmic selection criteria. These parameter changes enable the system to automatically identify and apply the most appropriate textures while maintaining high fidelity to the actual building façades.
3Productivity
If approximation methods are used to fill obstructed areas, then productivity increases, but manufacturing precision deteriorates due to loss of unique building features
Solution Approach 1:
The system extracts and removes occluded portions from the three-dimensional model or excludes them from texture application, rather than approximating them. By taking out the obstructed areas and only applying textures to visible, unobstructed portions, the system maintains accuracy of building features while improving productivity through automated processing of available imagery.
Data Source
AI summary
Computer systems and methods are described for automatically generating a 3D model, including, with computer processor(s), obtaining geo-referenced images representing the geographic location of a structure containing one or more real façade texture of the structure; locating a geographical position of real façade texture(s) of the structure; selecting base oblique image(s) from the images by analyzing image raster content of the real façade texture depicted in the images with selection logic; analyzing the real façade texture to locate a geographical position of at least one occlusion using pixel pattern recognition of the real façade texture; locating oblique image(s) having an unoccluded image characteristic of the occlusion in the real façade texture; applying the real façade texture to wire-frame data of the structure to create a 3D model of the structure; and applying the unoccluded image characteristic to the real façade texture to remove the occlusion from the real façade texture.