Location Management in a Structured Environment

Unmanned Aerial Vehicles navigation usually relies on precise localization services that are commonly acquired using a GPS system. However, the GPS signal can be lost or disturbed due to bad weather, the existence of obstacles or other constraints imposed by the environment. In this context, the Inertial Measurement Unit alone becomes ineffectual for providing reliable information about the vehicle position, because of the cumulative error on calculation iterations. Subsequently it produces mission errors, which could have disastrous consequences. This work presents as a contribution an alternative navigation strategy in the hovering framework, which makes the flying craft able to overcome the localization constraints at these environments. The proposed solution consists of a system that toggles from a GPS-based localization to hybrid localization. This hybrid approach is essentially based on two methods. The first will be based on deterministic method, justified by exact data that don????????t overlap position estimation, as against the second will be based on a Bayesian method to calculate the relative UAV position by the combination of the IMU data and the results of image processing results using the SIF descriptor, which is most appropriate for our context