Modelling camera viewing angle deviation to improve nearshore video monitoring

Clément Bouvier (1) , Yann Balouin (1) , Bruno Castelle (2) , Robert Holman (3)

1 BRGM - Bureau de Recherches Géologiques et Minières (BRGM)
2 EPOC - Environnements et Paléoenvironnements OCéaniques
3 CEOAS - College of Earth, Ocean and Atmospheric Sciences [Corvallis]

Video monitoring the nearshore can provide high-frequency remotely-sensed optical information from which morphological changes and hydrodynamic data can be derived. Although overlooked in most of the studies, it is acknowledged that camera viewing angles can substantially vary in time for a variety of reasons, reducing consistently data accuracy. This paper aims to identify the primary environmental parameters controlling camera shifts at the video monitoring station of Sète (SE France) and develops an empirical model to routinely reduce these deviations. Our model simulates camera movements with an excellent skill (BSS = 0.87) and shows that camera viewing angles’ deviation is primarily controlled by the position of the sun during sunny days, making it predictable. This study opens new perspective to routinely improve camera geometry of video monitoring systems.

Keywords: image rectification; Coastal management; video monitoring

Domains: Geomorphology; Ocean, Atmosphere

Clément Bouvier, Yann Balouin, Bruno Castelle, Robert Holman. Modelling camera viewing angle deviation to improve nearshore video monitoring. Coastal Engineering, 2019, 147, pp.99-106. ⟨10.1016/j.coastaleng.2019.02.009⟩. ⟨hal-02023192⟩