METHOD FOR SHIP DETECTION IN SYNTHETIC APERTURE RADAR IMAGERY BASED ON THE INTEGRATION OF IMAGE ENHANCEMENT AND ANALYSIS OF PIXEL VARIANCE

Abstract

In this study, a novel method to detect ships in the SAR imagery by the integration of image enhancement and analysis of pixel variance is suggested. The method was devised for the use on a specific module called ‘SAR-module’ which has been designed to provide a near-real-time ship inspection and detection using SAR images over a region of interest. For experiments, RADARSAT-2 HV data covering Ieodo ocean area of Korea is used. This work evaluates the level of enhancement using signal-to-noise ratio (SNR) and confirms that the proposed method significantly improved the SNR for all test area. This implies that the suggested enhancement method can be fully used to improve ship detection performance when integrating ship detection method based on analysis of pixel variance. We also confirm the potential of the method using the ship detection results. Overall, this work demonstrates the potential of the suggested method as an automatic ship detection method using SAR images for a near-real-time ocean surveillance.en designed to provide a near-real-time ship inspection and detection using SAR images over a region of interest. For experiments, RADARSAT-2 HV data covering Ieodo ocean area of Korea is used. This work evaluates the level of enhancement using signal-to-noise ratio (SNR) and confirms that the proposed method significantly improved the SNR for all test area. This implies that the suggested enhancement method can be fully used to improve ship detection performance when integrating ship detection method based on analysis of pixel variance. We also confirm the potential of the method using the ship detection results. Overall, this work demonstrates the potential of the suggested method as an automatic ship detection method using SAR images for a near-real-time ocean surveillance.