A Topology Map Generation Algorithm for Optimal Path Finding for Image-Based Maps

Abstract

Algorithms that automatically search for routes have various applications, such as games, navigation, and robots, and are used to create optimal paths in the space of real and virtual worlds. This paper introduces an indoor navigation algorithm allowing rescuers to create a topology map of floor plans, given that the amount of data provided by conventional mapping constructs is limited. Traditional pathfinding algorithms include the A* algorithm, which has demonstrated excellent performance; however, its applicability is limited in the case of high-resolution images, as the number of nodes increases significantly. Here, we use the corner node algorithm (CNA) to improve pathfinding performance by preprocessing the data to create a topology map for the pathfinding algorithm. The algorithm can reduce the number of nodes required for the pathfinding process by generating topology maps in advance through two preprocessing processes: node creation and node connection. This effectively reduces the number of nodes to be processed, as well as the error rate when creating more accurate routes. Furthermore, a generated topology map can be reused by the algorithm unless its structure changes, which enhances efficiency. For the user, this means that the sought-after mapping information of the floor plan is quickly made available for navigation. With this algorithm, we created an augmented reality navigation solution to display virtual objects for rescuers, to demonstrate the enhanced performance of our approach.