Effectual monitoring system for maritime transport is essential for every country military/navy security system. The demand is very high for security and commercial areas to detect unknown objects (or) ships. There are many reasons to implement this idea such as secured monitoring of ships in narrow canals, oil theft control, sea pollution monitoring, attacks of pirates, terrorism activities, identification of debris and lost ships, and identification of submarines. There are many ship detection methodologies which are available like vessel monitoring system, automatic identification system, and long-range identification and tracking, but these methods require very high complex hardware equipment, especially high-frequency transponders are required to find the ship location. With these traditional methods, there may be more chance of getting attacked by pirates (or) thieves. In this paper, we proposed a new fast learning method for ship detection by using satellite images (which captured by satellite). Deep learning models were used to implement secured ship detection, i.e., single shot detector (SSD). There are many methods like HOG, RCNN, and YOLO which were implemented by using machine learning algorithms, but quick and accurate detection is not possible in these methods. In our proposed method by using SSD, we may get accuracy up to 99%. Our modeling results will show that ship detection method for satellite images by using deep learning SSD model is best and accurate method.
Ship Detection from Satellite Images with Advanced Deep Learning Model (Single Shot Detector (SSD))
Smart Innovation, Systems and Technologies
International Conference on Frontiers of Intelligent Computing: Theory and Applications ; 2022 ; Aizawl, India June 18, 2022 - June 19, 2022
2023-02-24
13 pages
Article/Chapter (Book)
Electronic Resource
English
Automatic Vehicle Detection from Satellite Images Using Deep Learning Algorithm
Springer Verlag | 2021
|