Abstract
Synthetic aperture radar (SAR) images have been used in many studies for ship detection because they can be captured without being affected by time and weather. In recent years, the development of deep learning techniques has facilitated studies on ship detection in SAR images using deep learning techniques. However, because the noise from SAR images can negatively affect the learning of the deep learning model, it is necessary to reduce the noise through preprocessing. In this study, deep learning vessel detection was performed using preprocessed SAR images, and the effects of the preprocessing of the images on deep learning vessel detection were compared and analyzed. Through the preprocessing of SAR images, (1) intensity images, (2) decibel images, and (3) intensity difference and texture images were generated. The M2Det object detection model was used for the deep learning process and preprocessed SAR images. After the object detection model was trained, ship detection was performed using test images. The test results are presented in terms of precision, recall, and average precision (AP), which were 93.18%, 91.11%, and 89.78% for the intensity images, respectively, 94.16%, 94.16%, and 92.34% for the decibel images, respectively, and 97.40%, 94.94%, and 95.55% for the intensity difference and texture images, respectively. From the results, it can be found that the preprocessing of the SAR images can facilitate the deep learning process and improve the ship detection performance. The results of this study are expected to contribute to the development of deep learning-based ship detection techniques in SAR images in the future.
Highlights
Ship detection is a critical maritime management technology and encompasses fields such as the investigation of illegal fishing areas, oil spill detection, maritime traffic management, and national defense [1,2,3,4]
The synthetic aperture radar (SAR) image was preprocessed in three different ways to compare compare the effect on the ship detection result of different training data used during the deep learning the effect on the ship detection result of different training data used during the deep learning model model training
To further improve the ship detection performance, a preprocessing method was proposed to reduce the noise in the SAR image and increase the contrast between the sea and the ship
Summary
Ship detection is a critical maritime management technology and encompasses fields such as the investigation of illegal fishing areas, oil spill detection, maritime traffic management, and national defense [1,2,3,4]. If the density of a ship is high, the AIS signal may be affected by interference, which reduces the effective reception distance, and may not be received by other ships or base stations [5]. When AIS is deliberately turned off by a fishing ship, it is impossible to locate the ship. Satellite images can be used as an alternative for detecting ships in the sea. If satellite images are used, most ships in vast water bodies can be effectively detected regardless of the operation of the AIS. There are two main types of sensors for capturing satellite images: an optical sensor, which is a passive sensor, and a synthetic aperture radar (SAR) sensor, which is an active sensor. The optical sensor generates an image by detecting sunlight reflected from an object. Because the optical sensor uses the visible color spectrum to capture images, ships in images captured by an optical sensor can
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
