Building Scan Window Detection via Point Cloud Transparency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Older buildings often lack up-to-date 3D architectural models necessary for accurate energy consumption assessment due to undocumented remodeling, making it difficult to determine the number and size of windows, which are critical for energy consumption analysis.
Innovation Solution
A method using a building scan with an electronic computing device to detect windows by processing a colored point cloud, decomposing walls, extracting transparent features, classifying and smoothing them, and incorporating the detected windows into a 3D model, including steps like ray tracing, pixilation, and façade regularization to ensure accurate window detection and alignment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If building scan processing is used to detect windows in older buildings, then measurement precision of window detection is improved, but device complexity increases
Solution Approach 1:
The window detection process is divided into distinct sequential steps: receiving building scan data, processing to generate colored point cloud and 2D floor plan, decomposing point cloud into walls, rotating and pixelating walls, extracting transparent features, classifying and smoothing features, and incorporating windows into 3D model. This segmentation allows each step to be optimized independently while maintaining overall detection precision.
Solution Approach 2:
The patent introduces intermediate data structures including colored point cloud, 2D floor plan, and pixelated wall representations as mediators between the raw building scan and the final window detection. These intermediaries transform complex 3D scan data into formats suitable for transparent feature extraction, reducing the complexity of direct window detection from raw scan data.
2Manufacturing precision
If detailed processing of colored point cloud is performed to extract window features, then manufacturing precision of 3D model is improved, but loss of time increases
Solution Approach 1:
The patent performs preliminary processing steps before window extraction, including generating the colored point cloud and 2D floor plan from building scans, and decomposing the point cloud into wall structures. By preparing these intermediate representations in advance, the actual window feature extraction can focus on identifying transparent regions without reprocessing the entire scan data, thus reducing total processing time while maintaining precision.
Solution Approach 2:
The patent extracts only the relevant transparent features from the processed wall data that correspond to windows, rather than analyzing all points in the colored point cloud. This selective extraction of transparent features from the pre-processed wall representations reduces computation time while maintaining accurate window detection in the final 3D model.
3Measurement precision
If ray tracing is used to determine transparency of building scan portions, then measurement precision of window detection is improved, but use of energy increases
Solution Approach 1:
The patent applies ray tracing only to specific regions identified as potential window areas in the pixelated wall representations, rather than performing ray tracing on the entire building scan or all wall surfaces. By focusing computational resources on local regions with transparent features, the energy consumption is reduced while maintaining high precision in window detection where it is most needed.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables the creation of accurate 3D models of buildings, allowing for precise energy consumption analysis by effectively detecting and classifying windows, even in buildings with outdated or missing digital records, thereby improving energy efficiency assessments.
Implementation Method 1
detecting the windows based on no return transparency of portions of the building scan
Data Source
AI summary
Methods for detecting a windows in a building utilizing a building scan using an electronic computing device are presented, the methods including: causing the electronic computing device to receive the building scan; processing the building scan; detecting the windows based on no return transparency of portions of the building scan; and incorporating the detected windows into a 3D model corresponding with the building scan. In some embodiments, processing the building scan generates at least the following: a colored point cloud; a 2D floor plan of the building; and a walked path. In some embodiments, detecting the windows based on transparency of a portions of the building scan further includes: decomposing the colored point cloud into walls; rotating and pixilating the walls; extracting transparent features corresponding with the windows; and classifying and smoothing the transparent features corresponding with the windows.