Sector Window SAR Image Edge Detector With Edge Compensation Strategy
ShuJian Zhou, YaoZong Zhang, Xiaoming Li, Dan Zhao, Jiguang Mei · 2023
It is found that in the edge detection of synthetic aperture radar (SAR) images, conventional Gaussian-Gamma Shaped (GGS) and Ratio of Average (ROA) algorithms use detection windows that tend to cross over the true edges and contain regions of different quality, which leads to high false alarm rates. To solve this problem, the study proposes a new edge detector with a novel window design called the sector window (SW). This scalloped window allows better adaptation to the characteristics of SAR images, faster extraction of thin edges, and reduced inclusion area of different qualitative regions. In addition, in order to extract edges between similar uniform regions more accurately, the study also introduces an edge compensation strategy that enables the algorithm to detect some weak edges that are not easily extracted, thus improving the accuracy in the edge detection process. Both objective and subjective experiments show that the proposed edge detector has important potential for application in synthetic aperture radar image processing by introducing a new sector window and edge compensation strategy that can provide accurate edge detection results.