Localization in Aerial Imagery with Grid Maps using LocGAN
This addresses localization for autonomous vehicles in urban environments, but it is incremental as it builds on existing cGAN and localization network methods.
The paper tackles the problem of vehicle localization in urban areas where GNSS is unreliable and prior sensor-specific maps are unavailable, by using aerial imagery as prior and transforming it to match LiDAR grid maps via a cGAN, achieving reliable global localization results in evaluations on Karlsruhe data.
In this work, we present LocGAN, our localization approach based on a geo-referenced aerial imagery and LiDAR grid maps. Currently, most self-localization approaches relate the current sensor observations to a map generated from previously acquired data. Unfortunately, this data is not always available and the generated maps are usually sensor setup specific. Global Navigation Satellite Systems (GNSS) can overcome this problem. However, they are not always reliable especially in urban areas due to multi-path and shadowing effects. Since aerial imagery is usually available, we can use it as prior information. To match aerial images with grid maps, we use conditional Generative Adversarial Networks (cGANs) which transform aerial images to the grid map domain. The transformation between the predicted and measured grid map is estimated using a localization network (LocNet). Given the geo-referenced aerial image transformation the vehicle pose can be estimated. Evaluations performed on the data recorded in region Karlsruhe, Germany show that our LocGAN approach provides reliable global localization results.