Method for automatic georeferencing aerial remote sensing (RS) images from an unmanned aerial vehicle (UAV) platform

Before an aerial image can be used to support a site-specific application it is essential to perform the geometric corrections and geocoding. This research discusses the development of an automatic aerial image georeferencing method for an unmanned aerial vehicle (UAV) image data acquisition platform that does not require use of ground control points (GCP). An onboard navigation system is capable of providing continuous estimates of the position and attitude of the UAV. Based on a navigation data and a camera lens distortion model, the image collected by an onboard multispectral camera can be automatically georeferenced. When compared with 16 presurveyed ground reference points, image automatic georeferenced results indicated that position errors were less than 90 cm. A large field mosaic image can be generated according to the individual image georeferenced information. A 56.9 cm mosaic error was achieved. This accuracy is considered sufficient for most of the intended precision agriculture applications.

Research highlights
► Automatic aerial image georeferencing method developed for an unmanned aerial vehicle (UAV) image data acquisition platform.
► UAV images can be automatically georeferenced using a navigation data and a camera lens distortion model.
► Georeferenced errors are less than 90 cm.
► A large field mosaic image can be generated according to the individual image georeferenced information