摘要
传统的图像复原方法往往只考虑遥感图像获取过程中模糊和噪声对图像质量的影响,而忽略了图像混迭的影响.针对斜模采样系统调制传递函数各向异性的特点,通过自适应倒易晶胞确定图像中混迭和噪声较小的频谱覆盖范围;从图像复原的贝叶斯方法出发,以小波域隐马尔科夫树模型作为图像的先验模型,提出一种结合自适应倒易晶胞和隐马尔科夫模型的斜采样遥感图像复原方法.实验结果表明,该方法可有效地提高图像分辨率,较常用的图像复原方法复原效果更好.
T raditional methods of image restoration tend to only think of the influence of noise and blur on image quality during the image acquisition process , alias is ignored . Due to the anisotropic characteristics of modulation transfer function of the tilting mode sampling system ,the adaptation of the reciprocal cell can confine the shape of the unit with little alias and noise . In this paper , the Bayesian method and wavelet domain Hidden Markov Tree model are treated as prior models ,a method combining adaptation of the reciprocal cell with HM T model is proposed for tilting mode sampling image restoration .The experimental results show that the method can effectively improve the image resolution ,and is better than the commonly used image restoration methods .
出处
《计算机辅助设计与图形学学报》
EI
CSCD
北大核心
2014年第11期1966-1973,共8页
Journal of Computer-Aided Design & Computer Graphics
基金
国家科技支撑计划资助项目(N2012BAH31B01)
北京市自然科学基金重点项目(B类)(KZ201310028035)
