BUNDLE ADJUSTMENT ACCURACY ASSESSMENT OF UNORDERED AERIAL DATASET COLLECTED THROUGH KITE PLATFORM

Abstract. The introduction into the commercial market of affordable and off-the-shelves Unmanned Aerial Systems (UAS), have lately boosted the mapping capabilities of archaeologists. Hardware solutions have been indeed supported by more accurate flight planning software allowing to increase the reliability of 3D models in terms of spatial resolution and geometric accuracy. However, during the last decades, aerial photography was mainly performed exploiting imaging sensors mounted on kites, balloons and poles. Although being an affordable and user-friendly solution, the use of these platforms did not allow the collection of images following an ordered data collection, hence introducing factors in the network design which could hamper the photogrammetric reconstruction. This study aims to assess the Bundle Adjustment (BA) accuracy and the reliability of the photogrammetric reconstruction by reprocessing various dataset collected over the UNESCO archaeological site of Khirokitia Vouni (Cyprus) using a commercial software and DBAT (Damped Bundle Adjustment Toolbox).