中文摘要：

针对无源毫米波图像分辨率低的问题，文中提出一种将Richardson—Lucy（RL）迭代算法与Contourlet（CT）变换相结合的图像超分辨复原方法（Contourlet—RL，CT—RL）。该方法利用RL算法实现图像频谱外推，并采用Context模型对多幅迭代子图进行CT域自适应去噪，最后利用各子图CT域部分高频信息进行逆CT变换以获得高分辨率图像。经实验结果证明，该方法能有效改善RL算法在频谱外推过程中对噪声敏感的缺陷，增强了图像的分辨率，有利于无源毫米波图像的超分辨率复原。

英文摘要：

A new method named Contourlet-RL based on Richardson-Lucy algorithm and Contourlet transform is proposed for improving the poor resolution of passive millimeter wave （PMMW） images. The RL algorithm is used to spectral extrapolation ,meanwhile the noise of sub images is removed self-adaptively at the regional level of CT domain according to the model of Context, and then the part infor- mation of high frequency is extracted and assembled in CT domain. At last, the inverse transformation is executed for obtaining the re- stored image. Experimental results demonstrate that the method compensates for the flaw of the RL algorithm＇s noise susceptibility and en- hanee~ the t＇~gt~｜lltil[~rl ]zTiIvlharn-lnrt~ it i~ nmnltlnll~ ta th~ cnn~r--r~cnlntlnn r~ctor~tlnn ftxr PI~JlI~J[~ imanac