Application of NDWI and Machine Learning Techniques for Shoreline Extraction

Shoreline monitoring, stipulated as the approach allowing for the identification of the intersection between the land and water surfaces, is crucial for understanding and managing coastal environments. Remote sensing techniques, enhanced by advancements in deep machine learning and artificial intelligence, offer a promising alternative for shoreline monitoring. The article presents a method for coastline extraction using the NDWI index and Deep Machine Learning. Satellite images from Sentinel-2 were used for the study. These images covered the city beach in Gdynia and the adjacent waterbody. This area was chosen due to its frequent use as a measurement site in various environmental studies. Additionally, the article details the stages of the method and the tools used in Arc Pro software. The aim of this study is to develop an automated method for coastline determination that can serve as an alternative to traditional methods such as GNSS RTK receiver, LiDAR, or photogrammetry. The research findings indicate that the NDWI automatic method enables accurate coastline determination. Furthermore, the proposed automated coastline extraction method can complement field measurements. It may prove useful in bathymetric monitoring, where cyclical measurements of the coastline are required.