Mobile indoor positioning using Wi-Fi localization and image processing

In recent years, there has been an ongoing interest in indoor positioning systems. Many designs have been proposed which have employed a wide variety of algorithms, including Wi-Fi and image processing algorithms. Although current experiments have been designed that incorporates the use of individual algorithms, there is still much to account for, such as the lack of accuracy in Wi-Fi Localization techniques, and the lack of speed in image processing. In this study, a two-phase framework was designed to have one algorithm compensate for the other’s weakness. The algorithms used in this study were Wi-Fi Localization and image processing techniques. This framework implemented Wi-Fi Localization with routers in order to determine the user’s rough location, and applied image processing as a means to improve the accuracy of the predicted location. Techniques that involved image masking and low-resolution imagery were also integrated to improve image masking and low-resolution imagery were also integrated to improve speed without jeopardizing accuracy. Test have shown that the framework had better speed and accuracy as compared to using these algorithms individually, and it surpassed the accuracy of a number of current indoor positioning systems. Further analysis also allowed to determine the limitations of the framework, and suggestions were raised for additional refinement