摘要
相机成像过程中与地面的相对运动、遥感平台姿态变化、发动机振动、电磁波干扰等都会导致质退化而造成影像模糊,为此提出了一种基于二维经验模态分解算法的遥感影像去模糊方法。结合分解后的本征模函数IMF分量实现对影像的去噪和特征地物清晰边缘的提取,利用稀疏分解框架下的反卷积模型估计模糊核,通过带约束的最小二乘法进行影像解模糊。通过仿真实验验证了该方法的可行性。
The relative motion of the camera imaging process to the ground, attitude changing of the remote sensing platform, engine vibration, electromagnetic interference and others, all of these will affect the effect of the imaging result, lead to image quality degradation with blurring the image. This paper proposed a de-blur method for remote sensing images based on bidi- mensional empirical mode decomposition algorithm. Firstly, it achieved image de-noising and clear edge detection for surface feature with intrinsic mode functions, then estimated the fuzzy kernel combined clear edge and de-convolution model in the fame of sparse decomposition. It realized the de-blur result of fuzzy image with constrained least squares method. Finally, the result verifies the feasibility of the algorithm through simulation experiment.
出处
《计算机应用研究》
CSCD
北大核心
2015年第1期276-279,共4页
Application Research of Computers
基金
国家"973"计划资助项目(2012CB719901)
国家"863"计划资助项目(2012AA12A304)
国家自然科学基金资助项目(50848053)
关键词
遥感影像
退化模型
BEMD
点扩散函数
边缘检测
remote sensing images
degradation model
BEMD
point spread function
edge detection
