Performance Improvement for Indoor Positioning Systems Using Xtion Depth Sensors and Smartphone Orientation Sensors

Nowadays, smartphones and tablet computers are everywhere, and these mobile devices offer a lot of applications, such as the location-based services (LBS). The location-based services not only can provide the users’ locations, but also extend a lot of services, like navigation services and traffic information services, etc. Currently, the main stream of positioning technology is the global positioning system (GPS), but it is not suitable for indoor applications and services. The more suitable indoor positioning technology is the RF signal pattern matching method; whereas it will take a lot of time and man powers to build the RF map. In order to reduce the costs of time and man powers, this study proposes a system to build the indoor RF map through sensory detectors (ex. Xtion or Kinect) on the off-line stage. Through the capture of the depth information and the object image, we can build RF prediction propagation models from the environment information with less time and manpower cost. With built-in sensors (accelerometers and gyroscopes) of smart phones, we can estimate the users’ moving directions and distances in the environment with better positioning accuracy in online stage. Eventually, the paper cut down running time about 78% and obtains the double samples simultaneously. The positioning error is 2.73 meters indicated in the experimental results.