Automatic Floor Plan Analysis for Adaptive Indoor Wi-Fi Positioning System

摘要

An automatic floor plan analysis system is described which automatically extracts key geographic information from a floor plan and creates a weighted graph representation of the indoor environment. The purpose of this system is to support the Adaptive indoor Wi-Fi Positioning System (AWPS) we proposed earlier. The primary geographic information extracted includes corridor segments and their vertices, and room numbers of door entrances along corridor segments. We call the weighted graph with geographic information the Euclidean graph. The purpose of this Euclidean graph is two-fold. First, with automorphism detection, it serves as the key for associating the floor plan with the crowd-sourced radio graph. Second, it provides important geographic information for enriching indoor localization and navigation applications.