A RGB and D vision aided multi-sensor system for indoor mobile robot and pedestrian seamless navigation

An accurate navigation system is an essential and important part for many applications carried out in the indoor environments. In the absence of absolute positioning information such as global positioning system, the navigation solution which relies on previous system states such as dead reckoning has shown disadvantage over long time running. In contrast, we introduce an novel and inexpensive sensor fusion based approach to solve the problem. The key idea is to use visual gyroscope as a complementary source for system heading estimation since it only depends on the scene observed by the camera. The introduced methods only requires a Kinect and a low cost inertial measurement unit. By using the same mechanism, the introduced method has been tested for the applications of both indoor mobile robot and pedestrian navigation. Field experiments have been carried out and corresponding results are presented.