Resumen
The popularization of smart phones and the large-scale application of location-based services (e.g., exercises, traveling and food delivery via cycling) have resulted in the emergence of massive amounts of personalized cycling trajectory data, spurring the demand for map navigation based on cycling trajectories. Therefore, in the current paper, we propose a cycling trajectory-based navigation algorithm without the need for road network data support. The proposed algorithm focuses on extracting navigation information from a given trajectory and then guiding others to the destination along the original trajectory. In particular, the algorithm analyzes the coordinate and azimuth angle data collected by the built-in positioning and direction sensors of mobile smart phones to identify several turning modes from the provider?s cycling trajectory. In addition, the interference of the traffic conditions during data collection is considered in order to improve the recognition accuracy of the turning modes. The turning modes in the trajectory are subsequently transformed into navigation information and shared with users, so as to realize the shared navigation of the cycling trajectory. Experimental results indicate that the algorithm can accurately extract the turning feature points from cycling trajectory data, recognize various turning modes and generate correct navigation messages, thereby guiding users to arrive at the destination safely and accurately along the original trajectory. The algorithm is independent of electronic map platforms and does not require road network data support.