Joint use of image-based and GNSS techniques for urban navigation

Abstract

The use of position-based devices is constantly increasing with a wide spectrum of applications, e.g. the continuous demand of mapping services based on user’s location. Depending on the specific application, a different level of accuracy could be requested, going from room level to few centimeters of error. The navigation problem is typically faced by using Global Navigation Satellite Systems (GNSS), but this technique cannot efficiently work in case of poor sky visibility, as happens in urban areas. An option could be the combination of image-based and GNSS solutions. Two different assisted photogrammetry techniques are here presented and discussed. First, an image-based navigation solution constrained by using ground control points (GCPs) extracted from urban maps is presented. It was tested considering a set of different scenarios, reaching accuracies of the order of 0.20 m. A second outdoor navigation solution has been realized by integrating the data acquired by a Microsoft Kinect device (RGB and depth images) and a GNSS receiver through a proper Kalman filter. Also in this case the achieved accuracy is of the order of 0.20 m.