位置:成果数据库 > 期刊 > 期刊详情页
SAR图像点目标超分辨正则化方法的简化计算
  • 期刊名称:中国空间科学技术, Vol.28, No.1, 2008: 41-46
  • 时间:0
  • 分类:TN958[电子电信—信号与信息处理;电子电信—信息与通信工程] TP391.41[自动化与计算机技术—计算机应用技术;自动化与计算机技术—计算机科学与技术]
  • 作者机构:[1]国防科学技术大学理学院,长沙410073, [2]桂林空军学院,桂林541003
  • 相关基金:国家自然科学基金资助项目(No.60572136)
  • 相关项目:SAR图像数据的超完备稀疏表示及应用
中文摘要:

合成孔径雷达(SAR)是一种高分辨率相干成像雷达,是重要的空间遥感手段,分辨率的高低直接决定了其应用效果的优劣。正则化方法是提高SAR图像分辨率的有效方法,但传统的迭代算法计算速度较慢。文中分析了基于lk范数的图像超分辨正则化方法的迭代求解公式,得出了由该迭代公式产生的迭代序列收敛的极限的解析表达式。将该结论用于正则化模型的求解可以显著降低计算量,由此得到了正则化迭代算法的简化计算方法。仿真与MSTAR图像处理结果表明,简化后的方法计算速度提高了2000倍以上,计算结果与传统方法的相对差别小于6×10^-9。

英文摘要:

Synthetic aperture radar (SAR) which is a high resolution coherent radar is a significant space remote sensing method, and the resolution of SAR image determines its further application. Regularization method is effective to improve the resolution of SAR image, but the traditional iterative method for resolving the regularization model runs slow. The iterative formulation of image supper-resolution regularization model based on lk norm was analysed, The analytic expression of the limit of iterative sequence which is produced by the iterative formulation was obtained. According to the conclusion, a new approach to solve the regularization model was obtained, and it could significantly simplify calculation. Simulation and results of MSTAR image processing show that the efficiency of the calculation impr, oves more than 2000 times and the relative error between the presented method and the conventional method is only 6 × 10^-9.

同期刊论文项目
期刊论文 54 会议论文 2 著作 1
同项目期刊论文