Coastal monitoring and feature estimation with small format cameras: application to the shoreline of Monte Hermoso, Argentina

Abstract

Image and video processing of natural phenomena is one of the preferred non-invasive monitoring techniques for environmental studies that is, however, limited through the high cost of the required equipment and the limited access and precision of the processing algorithms. In this work we propose a low cost methodology for environmental studies using unexpensive off-the-shelf hardware and simple yet powerful processing algorithms. The images are taken using small format RGB cameras and processed in standard laptop equipments using open source libraries and processing algorithms specifically developed in general purpose programming languages. We applied this methodology to the coastal monitoring the shoreline of Monte Hermoso, Argentina, aimed at establishing accurate measurements of specific coastal features, for instance the coastal length. The experimental results show that our proposed unsupervised processing algorithm obtains results with a very high level of accuracy.