Georeferencing experiments with UAS imagery

Abstract. Comparing typical airborne mapping systems with Unmanned Airborne Systems (UAS) developed for mapping purposes, there are several advantages and disadvantages of both systems. The unquestionable benefits of UAS are the much lower costs of equipment and the simple operation; though, the regulations to fly UAS greatly vary by country. Low cost, however, means small sensor size and low weight, thus, sensors usually lack the quality, negatively impacting the accuracy of UAS data and, consequently, any derived mapping products. This work compares the performance of three different positioning approaches used for UAS image geolocation. The first one is based on using dual-frequency GPS data, post-processed in kinematic mode. The second approach uses the single frequency, code only GPS data that was acquired and processed by a geotagger, attached to mapping camera. Finally, the third one employs indirect image georeferencing, based on aerial triangulation using ground controls. As expected, the quality of data provided by the inexpensive GPS receiver (geotagger) is not suitable for mapping purposes. The two other approaches provided similar and reliable results, confirming that commonly used indirect georeferencing, which usually assures good solution, can be replaced by direct georeferencing. The latter technique results not only in reduction of field work, e.g. Ground Control Points (GCPs) surveying, but is appropriate for use with other sensors, such as active imaging technology, LiDAR, further extending UAS application potential.