Automated Floorplan Generation Using Depth-Optimized Edge Masks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing methods for generating floorplans of buildings are labor-intensive, costly, and time-consuming, involving manual sketching, measuring, and drafting.
Innovation Solution
A method using a computer to receive image and depth information of a building, generate an edge existence probability mask, optimize it using depth information, derive semantic information, and create a pictorial rendering of the building, including wall edges.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual sketching and measuring methods are used to generate floorplans, then detailed and accurate floorplan information can be obtained, but the process requires significant manual labor, time, and cost
Solution Approach 1:
The patent replaces the manual mechanical process of sketching, measuring, and drafting with an automated computer-based system that captures images, processes depth information, and generates floorplans algorithmically. This substitution eliminates manual labor while maintaining measurement precision through automated edge detection and semantic analysis.
Solution Approach 2:
The system creates a digital copy of the physical building space by capturing images and depth information, then processes this digital representation to generate an accurate floorplan. This copying approach allows rapid reproduction of floorplans without repeated manual surveying.
2Loss of information
If manual surveying and drafting procedures are employed, then comprehensive building layout information can be captured, but the process is time-consuming and costly
Solution Approach 1:
The system performs continuous image capture and processing as the camera moves through the building space, allowing comprehensive data collection without interrupting the surveying process. The automated pipeline continuously processes images and depth information to generate complete floorplan data rapidly.
Solution Approach 2:
The system performs preliminary processing of images and depth information to identify edges, walls, and semantic features before final floorplan generation. This preliminary analysis ensures comprehensive information capture while streamlining the subsequent rendering process.
3Productivity
If automated image processing is used to generate floorplans, then manual labor and time are reduced, but challenges arise in accurately identifying wall edges and semantic information
Solution Approach 1:
The patent combines multiple data sources including images, depth information, edge existence probability masks, and semantic information to identify wall edges. This merging of multiple information streams compensates for individual method limitations and improves edge identification accuracy while maintaining automated efficiency.
Solution Approach 2:
The system uses recursive procedures that incorporate feedback from depth information to optimize edge existence probability masks. This iterative refinement process continuously improves wall edge identification accuracy based on feedback from multiple processing stages.
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
This method automates the generation of floorplans, reducing manual labor, costs, and time, while providing accurate and efficient pictorial renderings of building layouts.
Implementation Method 1
receiving, by the computer, from a light detection and ranging (LiDAR) device, depth information associated with the at least the portion of the building
Data Source
AI summary
Image data and depth information of at least a portion of a building are received. An edge existence probability mask is generated based on the image data. The edge existence probability mask represents confidence levels associated with identifying each of the one or more edges of a surface in the at least the portion of the building. The edge existence probability mask is optimized using the depth information to obtain an optimized edge existence probability mask. The one or more edges of the surface are then identified based on the optimized edge existence probability mask.