A database of hand-captured shorelines to support A.I. based detection algorithms.

One of the most important aspects of coastal analysis and management is the detection of coastlines, which has a wide range of applications, including assessments of hazards and environmental monitoring. The building of a database consisting of hand-captured shorelines is presented in this thesis. The purpose is to support artificial intelligence-based detection algorithms in shoreline analysis. To begin, the research investigates the fundamental notion of shorelines and then proceeds to perform a literature review to contextualize the significance of the study under investigation. Several technical elements, including the camera model, more specifically the pinhole camera and the collinearity equation, are investigated to provide a mathematical model of digital images and perspective changes which are implemented in the Python's OpenCV package used in the TAO project. Using a graphical comparison to highlight the differences between the two methods, the thesis then goes on to evaluate the hand-made shoreline detection method in contrast to the TAO-made shoreline detection approach. The research shows the advantages and limits of each strategy using this comparative analysis. It also emphasizes the significance of utilizing hand-captured shorelines as a possible approach to calibrate artificial intelligence-based detection algorithms for coastal monitoring and analysis.