Applicability of an Image-Based Estimation Method of Nearshore Morphology Using Small Unmanned Aerial Vehicle

Abstract

Systematic monitoring of nearshore area provides useful information on sandy beaches over a wide range of temporal and spatial scales. In this study, accordingly, a simple local remote sensing system is developed to monitor the morphological variations of sandy beaches. This monitoring system consists of acquisition of geo-referenced video image of nearshore area from a small UAV (Unmanned Aerial Vehicle) and subsequent image analysis. Owing to the rapid development of information technology, high resolution photographic images of sea surface can be easily recorded at favorable location in a cost-efficient way. The subsequent quantification of morphological changes is carried out based on bright intensity patterns. First, the video images are converted to successive snapshots and rectified. After removing the small oscillations through semi-automatic identification of Ground Control Points (GCPs), the time-stack images of bright intensity variations are constructed for a series of cross-shore sections located at specified alongshore intervals. For each cross-section, the crest lines of waves are tracked out by inspecting the location of steep gradient in bright intensity variations. The local tracking results provide the celerity of waves. Combined with the observed wave period, the local water depth is estimated based on the linear dispersion relation. The system has been applied to the field observation of Uchinada Coast, Ishikawa, Japan facing to the Sea of Japan. The accuracy of geo-referencing was shown to be as small as a couple of pixels. The accuracy of morphological estimation based on image processing has been confirmed through comparison with a field survey using a jet bike. The image-based estimation results qualitatively reproduced the patterns of morphological variation. The typical error was in the range 0.2 to 0.8 m. These results demonstrated the capability of the developed system to remotely estimate the coastal morphology on sandy beaches.