Glint Correction of Unmanned Aerial System Imagery

Glint in aquatic imagery captured by Unmanned Aerial Systems (UAS) is a limiting factor when performing spectral analysis. It cannot be corrected by methods developed for space-based imaging systems, meaning new approaches are required. Two processes using in-situ radiometric data were developed augmenting an established method for removing atmospheric effects from imagery, the Empirical Line Method (ELM), to remove glint from multispectral UAS imagery. The results of this correction showed good agreement with in-situ spectroradiometer measurements and similar accuracy to atmospherically compensated satellite measurements. The Root-Mean-Square Error of the UAS retrieved remote sensing reflectance was as low as 0.0004 sr(-1) and outperformed the traditional ELM.