Vision-based monitoring of heritage monuments: Unmanned Aerial Systems (UAS) for detailed inspection and high-accuracy survey of structures

Abstract

The maintenance and preservation of heritage structures is of particular importance from a historical, cultural, socio-political, economic and not least architectural and aesthetic point of view. An efficient maintenance scheme requires periodical inspections associated with a reliable condition determination of the structure. Especially the process of data acquisition is a time consuming, labour intensive and technically complex task. This paper presents a vision-based inspection and monitoring approach for heritage structures based on high quality aerial photos taken by remote controlled Unmanned Aerial Systems (UAS). In terms of quality, time and investment this approach provides an important contribution to data acquisition processes and monitoring strategies and will provide an increased efficiency in preserving architectural heritages. UASs equipped with high quality photo or video cameras open up the possibility of simplifying and accelerating data acquisition and monitoring tasks. The extreme flight stability of the UAS and the fully compensated camera provide high quality data which are essential for reliable condition monitoring and inspection results. This offers also the possibility of highly accurate survey and geo-referenced 3D-modelling of structures in relation to direct referenced visualisation of detected damages. The paper shows three case studies. Furthermore, it presents a combination of conventional inspection measurements with modern photogrammetric and computer vision methods for an automatic post-flight geo-referenced 3D reconstruction based on aerial images.