Iceberg Detection in Satellite Images using Deep Learning Techniques

Abstract

Iceberg detection is found to be more critical in the previous researchers. High quality satellite monitoring of dangerous ice formations is critical to navigation safety and economic activity in the regions. The satellite images play a crucial role in the identification of the icebergs. In this manuscript, a convolutional neural network (CNN) model is proposed for the iceberg detection from the satellite images. It is based on the satellite dataset for target classification and target identification. The iceberg detection is based on the statistical criteria for finding the satellite images. This model is used to identify automatically whether it is remote sensed target is iceberg or not. Sometimes the iceberg is wrongly classified as ship. This model is done to make accurate about the changes in the detection.