HWVaP: A smartphone-based hierarchical WiFi/vision assisted PDR fusion system for accurate indoor positioning

Abstract Smartphones have become the primary platform for indoor location-based services (LBS). Although pedestrian dead reckoning (PDR) alleviates some issues of pure inertial navigation, it is affected by cumulative errors over time. Vision-based positioning, though less susceptible to noise and infrastructure-dependent, often faces a trade-off between the high accuracy of image feature matching and its substantial computational cost, making it unsuitable for real-time applications. To address these challenges, this paper proposes an integrated WiFi/Vision/PDR indoor positioning method. First, we introduce a WiFi/vision fusion absolute positioning method. This approach utilizes WiFi fingerprinting and the smartphone's heading angle to constrain the image-matching scope, significantly improving both the speed and accuracy of image retrieval. Furthermore, a multi-matching image spatial intersection method is proposed…