摘要
提出一种基于参数活动轮廓模型的多模型融合的合成孔径雷达SAR(Synthetic Aperture Radar)图像目标轮廓提取方法,即在活动轮廓模型Balloon中引入新兴统计分布模型G0分布、基于区域的统计活动轮廓模型和多边缘检测算子模型,获得了一种新的目标轮廓提取方法。基于MSTAR项目的真实SAR图像的实验结果表明,本文所提出的方法能准确地获得SAR图像目标轮廓,可用于执行实际的SAR图像轮廓提取任务,为后续的SAR图像自动识别和特征级图像融合等任务提供了较为优良的输入信息。
This paper proposes a method of target contour extraction based on active contour model and multi-model fusion. On the basis of Balloon model, an improved active contour model, we incorporate G0 statistical distribution, regional statistics based active contour model, and multi-edge detection operator model, to obtain a new target contour extraction method. The proposed method is applied to the real SAR image from MSTAR program, and the results show that the target contour is accurately extracted. So, the proposed method may be used in SAR image interpretation for the task of extracting target contours to provide excellent input information for the subsequent tasks in SAR image interpretation like automatic identification and feature-level image fusion, etc.
出处
《电子技术应用》
北大核心
2013年第9期85-88,共4页
Application of Electronic Technique
基金
国家自然科学基金项目(61102138
61074161)
关键词
合成孔径雷达
目标轮廓提取
活动轮廓模型
多模型融合
G0分布
synthetic aperture radar
target contour extraction
active contour model
multi-model fusion
G0 distribution
