Algorithm for Shoreline Position Determination in a Dynamic Coastal Environment Using Images Obtained from a Video Monitoring System

Video systems are extensively utilized globally in coastal monitoring strategies, facilitating high temporal and spatial sampling frequencies with reduced logistical efforts. This paper presents a novel and simple tool for processing coastal images, designed for automatic shoreline detection and data analysis using MATLAB. The algorithm is based on applying the region-grow function to pre-selected points and extracting the border between these points which corresponds to the sea/land boundary from averaged images produced from a Video Monitoring System. The study was conducted at Hellshire Bay Beach on the southern coast of Jamaica. The variance/error of the shoreline position produced by the algorithm when compared to a manually delineated shoreline is on average -1.09 and +1.11 pixels, which is comparable to other studies. The data from a single day within the period between September, 2022 and February, 2023 illustrate that the Hellshire beach is highly dynamic, with both erosion and accretion occurring in distinct sections of the beach, corroborating findings from other studies in the area. This further validates the accuracy of both the images generated by the video monitoring system and the algorithm developed to delineate shoreline positions from these images.