Robotic construction maps matched to floor plans despite scale and gaps
Reading the Room: Scale-ambiguous Semantic Registration of Sparse VSLAM Maps to Floor Plans in Partially Built Environments
RoboticsComputer Vision and Pattern Recognition
Summary
Robots used in construction need to match their camera-based maps to building floor plans, but this is hard because the maps don't have exact sizes and construction sites are often incomplete and look very similar. The authors created a new method called Reading the Room that finds a good match by combining semantic (meaningful features) and geometric information, even when the scale is unknown or the site is only partially built. Their approach works better than existing methods, improving accuracy significantly. This helps robots understand where they are inside buildings under construction.
What this means in practice
- •For construction robotics teams: Register robot sensor paths to building plans for improved progress tracking on active construction sites with partial structures.
- •For indoor mapping service providers: Align visual mapping data from mobile devices to architectural floor plans despite scale uncertainty and incomplete environments.
Authors
Bjorn Stragier, Emilia Szymańska, J. Micah Prendergast, Lukas Rosenberger Schmid
Abstract
Many robotic construction applications, such as progress tracking, require registering a sensor trajectory to a prior 2D floor plan. Visual capture is attractive for its low cost and availability, but aligning a visual simultaneous localization and mapping (VSLAM) map to a floor plan is difficult; VSLAM recovers geometry only up to an unknown similarity transform, the sparse point cloud is cross-modal to the symbolic floor plan, sites are only partially built during capture, and construction sites are visually repetitive, degrading map quality while removing distinctive features. Existing methods address at most two of these challenges. We thus introduce Reading the Room(RR), a global-local approach to register scale-ambiguous, sparse, semantic VSLAM maps to partially built floor plans. Our key idea is to define a single robust, differentiable, class-balanced semantic-geometric energy function that governs all stages of the pipeline. We develop two global initialization schemes, leveraging a single inaccurate user-provided correspondence if provided, or estimating the initialization globally using a cross-modal matching network. Both schemes are designed to find initial candidates with respect to the energy function, which is then directly optimized for local refinement. We evaluate RR on 33 sequences across three active construction sites against classical and learning-based baselines, achieving a 4.5x improvement in mean absolute trajectory error over the strongest baseline in the general unconstrained setting, which increases to 7.6x when an anchor is provided.