基于PDE与小波的图像放大研究  

Image Zooming Based on PDE and Wavelet

作　　者：戴存兵 高永光 刘建平 何鹏程 吴绍民[2] 

机构地区：[1]61683部队 [2]信息工程大学科研部

出　　处：《信息工程大学学报》2012年第1期82-85,共4页Journal of Information Engineering University

摘　　要：提出了一种基于PDE(Partial Differential Equations)和小波插值的遥感图像放大方法。首先采用双三次插值方法,进行遥感图像的初始放大,再进行小波分解,将低频分量替换为原图像,小波合成,获得小波放大结果,并采用PDE方法中的自蛇模型做后处理。同经典的图像放大方法进行对比实验表明,文章方法能够有效地解决遥感图像放大后边缘模糊的问题,能够更清晰地显示地面细节,更好地显示地物目标的轮廓。This paper brings forward a method of remote sensing image zooming based on PDE and wavelet interpolation. First, bi-cubic interpolation is adopted for initial remote sensing image zooming. Then low-frequency components are replaced by original images with wavelet decomposition. The results of wavelet zooming after wavelet synthesization are processed with the self-snake model based on PDE. Compared with typical image zooming,this method can effectively solve the problem of edge fuzzy of zoomed remote sensing images, with ground details and object profiles shown more clearly.

关 键 词：图像放大 小波分析 偏微分方程 自蛇模型 

分 类 号：TN911.73[电子电信—通信与信息系统]