.png){kind=icon}
返回
A Segmentation-Based CFAR Detection Algorithm Using Truncated Statistics
.png){kind=icon}
OA
AI
DOI：10.1109/TGRS.2015.2506822.png)
.png){kind=icon}
.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
Target detection in nonhomogeneous sea clutter environments is a complex and challenging task due to the capture effect from interfering outliers and the clutter edge effect from background intensity transitions. For synthetic aperture radar (SAR) measurements, those issues are commonly caused by multiple targets and meteorological and oceanographic phenomena, respectively. This paper proposes a segmentation-based constant false-alarm rate (CFAR) detection algorithm using truncated statistics (TS) for multilooked intensity (MLI) SAR imagery, which simultaneously addresses both issues. From our previous work, TS is a useful tool when the region of interest (ROI) is contaminated by multiple nonclutter pixels. Within each ROI confined by the reference window, the proposed scheme implements an automatic image segmentation algorithm, which performs a finite mixture model estimation with a modified expectation-maximization algorithm. Data truncation is applied here to exclude all possible statistically interfering classes, and sample modeling is based upon the truncated two-parameter gamma model. Next, CFAR detection is conducted pixel by pixel, utilizing the statistical information obtained from the segmentation process within the local reference window. The segmentation-based CFAR detection scheme is examined with real Radarsat-2 MLI SAR imagery. Compared with the conventional CFAR detection approaches, our proposal provides improved background clutter modeling and robust detection performance in nonhomogeneous clutter environments.
Keyword:
Constant false-alarm rate (CFAR)
expectation-maximization (EM) algorithm
modeling
nonhomogeneous
sea clutter
segmentation
synthetic aperture radar (SAR)
target detection
truncated statistics (TS)
AI总结
对已上传原文的论文进行重点信息的提取，主要内容包括：简要概述、研究摘要、背景介绍、关键亮点、图文解析、展望与总结。
期刊
IF:
8.6
论文数:
2.1W
被引数:
10.7W
机构
引用论文
Crystal Structure of Dimeric Heme Oxygenase-2 from Synechocystis sp. PCC 6803 in Complex with Heme,.png)
Biochemistry
BiochemistryIF0
PREAI
分享收藏
查看更多内容