Abstract
For wave disaster mitigation, it is essential to collect and analyze a wide range and long-term continuous information of beach monitoring, for which the required cost and time can often be enormous. On the data collection stage, the authors have already established an inexpensive local remote sensing method to obtain continuous beach images from a network camera. In this study, accordingly, focus was placed on the improvement of the efficiency and availability of the shoreline extraction in the image processing. In the previous studies, binarization analysis based on the threshold of luminance value or manual processing with digitizing device have frequently been used as a method of the shoreline identification. The former lacks applicability and stability in the field, and the latter has drawbacks in efficiency. In order to improve the efficiency as well as to extend the range of applications, this study proposes a novel image analysis method that can automatically extract the position of the shoreline under a wide range of conditions. The applicability to the analysis of waves and currents has also been examined.
