Monitoring of Irregularity on Sea Surface from Land-Taken Images

Abstract

Monitoring of irregularities that will occur on the sea surface is important for environmental safety. In this study, photographs taken from the land were used to determine the irregularities on sea surface. The fact that fixed cameras placed on land and scanning the sea surface were installed in certain regions and the possibility of installing new ones has been an incentive for this study. In the study, the images created by the mucilage experienced in the Marmara Sea at the beginning of 2021 were used as the irregularity data. The authors developed an original model for this study based on traditional image processing methods. In addition, classification was made with a convolutional neural network-based method and the results were compared.