Graph-Based Floorplan Reconstruction for 3D Modeling

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current methods for reconstructing indoor three-dimensional models, particularly floorplans, face challenges due to severe illumination variability, large untextured areas, and clutter, leading to inaccurate and overly complex reconstructions with issues like shrinkage bias and sensitivity to initialization problems.

Innovation Solution

A computer-implemented method that estimates floor and ceiling heights, identifies a core region in a two-dimensional graph, and determines a solution path that minimizes a cost formula to generate a three-dimensional model of a building's interior, using a multi-view stereo algorithm and texture mapping for photorealistic results.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If traditional volumetric segmentation methods are used for floorplan reconstruction, then the method can handle noisy data, but it exhibits shrinkage bias causing thin walls to fail recovery

Engineering Contradiction:

Improverobustness to noisy dataVSAvoidthin wall recovery accuracy

Core Design Contradiction:

ReliabilityVSManufacturing precision

Solution Approach 1:

The patent changes the fundamental parameter of the reconstruction approach from volumetric segmentation to graph-based shortest path optimization. By representing the floorplan as a graph where nodes are image pixels and edges represent potential wall segments, the method transforms the continuous volumetric problem into a discrete path-finding problem that naturally preserves thin wall structures without shrinkage bias.

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The patent replaces the mechanical volumetric segmentation process with an information-based shortest path algorithm. Instead of iteratively categorizing voxels as indoor or outdoor, the method uses cost minimization along paths in a graph to determine wall locations, substituting a computational information processing approach for the traditional spatial partitioning mechanism.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Stability of the object's composition

If regularization terms are applied in volumetric segmentation, then the method can stabilize the solution, but it creates difficulty with local minima and sensitivity to initialization

Engineering Contradiction:

Improvesolution stabilityVSAvoidinitialization sensitivity

Core Design Contradiction:

Stability of the object's compositionVSDevice complexity

Solution Approach 1:

The patent inverts the traditional approach by instead of adding regularization terms to stabilize volumetric segmentation, it formulates the problem as finding the shortest path in a graph. This inversion transforms a constrained optimization problem prone to local minima into a path-finding problem where the global optimum can be found through dynamic programming or Dijkstra's algorithm, eliminating initialization sensitivity.

Inventive Principle:

Principle #13The other way round (Inversion)

3Ease of manufacture

If basic laser ranging systems are used for three-dimensional reconstruction, then the system is simple to implement, but the reconstruction accuracy is inferior

Engineering Contradiction:

Improveimplementation simplicityVSAvoidreconstruction accuracy

Core Design Contradiction:

Ease of manufactureVSMeasurement precision

Solution Approach 1:

The patent introduces an intermediary computational process between image capture and three-dimensional model generation. Instead of directly converting images to 3D models through simple geometric processing, the method uses graph-based shortest path optimization as an intermediary step to first determine the floorplan structure, which then guides the generation of the accurate three-dimensional model, achieving high precision without complex hardware.

Inventive Principle:

Principle #24Intermediary (Mediator)

4Productivity

If point matching and camera calibration are used for indoor scene reconstruction, then the method can process the data, but errors are easily introduced due to illumination variability and untextured areas

Engineering Contradiction:

Improvedata processing capabilityVSAvoidpoint matching accuracy

Core Design Contradiction:

ProductivityVSMeasurement precision

Solution Approach 1:

The patent extracts and removes the problematic point matching and camera calibration steps from the reconstruction pipeline. By formulating the problem directly in terms of finding the shortest path through image space to represent walls, the method eliminates the need for feature point matching and camera calibration, thereby removing the source of errors related to illumination variability and untextured areas while maintaining data processing capability.

Inventive Principle:

Principle #2Taking out (Extraction)

Data Source

PatentUS9025861B2System and method for floorplan reconstruction and three-dimensional modeling

Publication Date: 2015.05.05 GOOGLE LLC

AI summary

Systems and methods for reconstructing a floorplan of a building for generating a three-dimensional model are provided. One aspect of the present disclosure is directed to a computer-implemented method for generating a three-dimensional model of a building. The method includes estimating a floor height and a ceiling height of the building. The method also includes identifying a core region of a two-dimensional graph, the core region corresponding to an interior of the building. The method includes determining a solution path that circumnavigates the core region and minimizes a cost formula, the cost formula providing an edge cost for each of a plurality of edges. The method further includes generating a three-dimensional model of the interior of the building based on the floor height, the ceiling height, and the solution path.