3D Layout Model Generation from 2D Images with Boundary Tolerance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The construction of 3D room layouts using RGBD cameras or LIDAR scanners is costly due to the expensive hardware required, necessitating a more affordable and accurate method.
Innovation Solution
A method and system that generate a 3D layout model using digital images, employing instance segmentation, depth estimation, and boundary tolerance data to construct planes and intersections, allowing for the creation of professional-grade 3D layouts without the need for expensive depth sensors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If RGBD cameras or LIDAR scanners are used to construct 3D room layouts, then measurement precision is improved, but device cost increases
Solution Approach 1:
The patent creates a digital copy of the 3D environment by generating a 3D layout model from 2D images. Instead of directly measuring with expensive depth sensors, the system captures 2D images and synthesizes depth information through computational methods, effectively copying the spatial structure digitally to achieve accurate 3D reconstruction without costly hardware
Solution Approach 2:
The patent replaces the mechanical/optical measurement system (RGBD cameras and LIDAR scanners) with a computational image processing system. By substituting physical depth sensing mechanisms with algorithms that infer 3D structure from 2D images, the system eliminates the need for expensive depth-sensing hardware while maintaining measurement capability
2Device complexity
If computational methods are used to generate depth data from 2D images, then device cost is reduced, but measurement precision may deteriorate
Solution Approach 1:
The patent applies segmentation by dividing the image processing into distinct components: instance segmentation to identify individual objects, boundary detection to locate edges, and plane segmentation to separate different surfaces. This segmented approach allows each component to be optimized independently, improving overall depth estimation accuracy from 2D images
Solution Approach 2:
The patent introduces boundary estimate data and boundary tolerance data as intermediary elements between 2D images and 3D depth reconstruction. These intermediaries provide structured constraints that guide the depth estimation process, ensuring that the computed 3D layout adheres to geometric consistency and physical plausibility, thereby maintaining measurement precision
3Manufacturing precision
If boundary tolerance data is used to create plane buffers, then manufacturing precision is improved, but data processing complexity increases
Solution Approach 1:
The patent applies beforehand cushioning by creating plane buffers that extend boundary estimates by a predetermined distance before final 3D model construction. This preliminary expansion creates a tolerance zone that anticipates and compensates for potential boundary detection errors, ensuring robust and accurate plane definitions in the final 3D layout model
Data Source
AI summary
A computer-implemented method and system relate to generating a three-dimensional (3D) layout model. Segmentation masks are generated using a digital image. The segmentation masks identify architectural elements in the digital image. Depth data is generated for each segmentation mask. A set of planes is generated using the depth data and the segmentation masks. Boundary estimate data is generated for the set of planes using boundary data of the segmentation masks. A set of plane segments is generated by bounding the set of planes using the boundary estimate data. Boundary tolerance data is generated for each boundary estimate data. A 3D layout model is constructed by generating at least a boundary segment that connects a first bounded plane and a second bounded plane at an intersection, which is located using the boundary estimate data and the boundary tolerance data.