An Adaptive Map-Matching based on Dynamic Time Warping for Pedestrian Positioning using Network Map

This study aims to recognize a pedestrian's position in a metropolitan environment. In this study, we attempt to improve the accuracy of the positioning result by combining a map-matching method with the dead-reckoning estimation result estimated using portable sensor values. We use a network map for map-matching and apply the Dynamic Time Warping algorithm as the matching method to realize a robust map-matching for a pedestrian's complex walking route pattern. In an experiment in the city of Ginza, Tokyo, which covers an area of 800 square meters, the duration for one test case ranged from 30 min to 1 h. According to our evaluation, our proposed method has an average positioning error of 2.9 m Therefore, the proposed method has robust positioning in case of variable pace walking, meandering, and random stoppages.