Adaptive multi-image direct georeferencing for unmanned aerial vehicle application

Due to the flexibility and unmanned property, unmanned aerial vehicle (UAV) has been used on a wide range of applications and conditions recently. In the large number of possible applications, real time and accurate georeferencing is very important. Direct georeferencing with out ground control points is defined as the direct measurement of the imaging sensor external orientation parameters (EOP) using positioning and orientation sensors, which is a popular research area. This paper proposes a new adaptive multi-image direct georeferencing for UAV applications. Based on bundle block adjustment, the proposed method designs an adaptive initialization strategy for different image data and application scenes. Due to the error cancellation property of multi directional images, it improves the localization accuracy. The experiments show exciting results.