An Efficient Wi-Fi–Vision Map Construction and Self-Maintenance Method for Indoor Localization

Currently, WIFI-based indoor localization methods have been proven to be promising due to their low deployment cost. However, the overhead of constructing and maintaining maps remains a bottleneck for the widespread deployment of WIFI-based indoor localization methods. In this paper, we propose a novel combined WIFI and vision to construct and maintain maps. This method consists of three parts (including constructing the logarithmic distance path loss (LDPL) model using an improved whale optimization algorithm (IWOA), a novel fusion localization module (called LDPL-PF), and a light-weight threshold-based map maintenance model). Specifically, the LDPL model based on IWOA can first construct high-quality maps with limited data. Then, LDPL-PF localization method is used to determine the user’s location based on the map. Finally, a feedback mechanism is introduced to achieve map maintenance automatically. The localization results and the collected data are fed into a multi-decision mechanism to build a feedback network for long-term map maintenance. Extensive experimental results show that our proposed method has good accuracy and stability with state-of-the-art methods.