Rockfall detection along road networks using close range photogrammetry

Early warning for rockfall incidents occurring on slopes along highways and roads is a vital safety practice for both human lives and property loss. Monitoring systems mainly rely on laser scanners, distributed sensors and precision geodetic measurements. In this study, close range terrestrial photogrammetry is evaluated towards rockfall incident detection. Photogrammetry has been extensively used for 3D mapping and reconstruction of terrain and infrastructure mainly due to its relatively low implementation cost, compared to the above mentioned approaches, especially when using non-metric digital cameras. Yet, the underlying problems in terrestrial photogrammetry include processing time, precision, night capturing limitations, and various errors, including occlusion, camera geometry, lighting, etc. In the present paper, an experimental setup along with preliminary results of the terrain photogrammetry integration into an operational slope monitoring scheme are presented. More specifically, assessment of the detection limitations are provided in order to evaluate whether photogrammetry can constitute a cost-effective alternative for rapid surface and change detection processes. In addition, since false alarm incidents cause extensive resource draining, their magnitude and occurrence are also reported.