Deep Learning Framework for Edge Detecting and Classification in SAR Images of Oil Spills using CNN

Oil spills are a type of pollution that occurs when small or large quantities of crude oil or distilled oils, such as gasoline or diesel, pollute bodies of water and coastal lands. The spill contains toxic chemicals that can have serious short- and long-term health consequences for oil workers, nearby residents, marine life, animals, and the environment. There are numerous types of oil spills, the majority of which are minor, such as when oil spills from a ship while it is being refueled. Serious incidents, on the other hand, have long-term consequences. The most of major oil spills occur when a pipeline ruptures, a tanker sinks or runs aground, or when a drilling operation goes wrong. An oil spill is an untidy and risky occurrence that can harm people and animals in the surrounding area. The proposed work uses a Deep Learning model named MASK RCNN algorithm (MASK Region Based Convolutional Neural Network) to automatically detect the area or size of the oil spill to overcome the hazardous situation. Convolutional neural networks, a subset of artificial intelligence (AI), are trained on a visual dataset of oil spill images from various altitudes and geographical locations in this study. Experiments show that the proposed algorithm has a good ability to detect oil spills at a higher detection rate, with an identification rate greater than 80% and a false alarm rate less than 12%. The evaluation results show that the proposed algorithm performs well in detecting oil spills, with an overall detection rate of more than 80%.