摘要
提出了一种基于Contourlet变换的空间域和变换域增强相结合的遥感图像增强算法。首先对原图像进行拉普拉斯塔式变换(LP)得到原图像的细节图像,并将它和原图像线性相加实现空间域增强;然后对空间域增强后的图像进行Contourlet变换得到不同尺度和不同方向上的变换系数,构造非线性增强函数对变换系数做增强处理,实现变换域增强;最后对增强处理后的变换系数进行Contourlet反变换,实现最终的图像增强。实验结果显示:与其他增强算法相比,用本文算法增强后图像的均值可达到127,信息熵提高了10%以上,清晰度提高了7%以上,表明遥感图像得到了更好的增强效果。
In order to enhance the contrast and details of remote sensing image, an algorithm based on Contourlet transform for remote sensing image enhancement both in space domain and transform domain is proposed. Firstly, the linear addition of original image and its details obtained by Laplacian pyramid transform is used to enhance image in space domain. Then, the coefficients in different scales and different directions are obtained by Contourlet transform and a nonlinear function is introduced to modify the coefficients and enhance image in transform domain. Finally, the image enhancement is implemented by reconstructing these coefficients. The experimental results show that the mean value of enhanced image by this algorithm is about 127; the entropy and definition can be increased by 10% and 7% respectively compared with those of other algorithms. So this algorithm can get better effect for remote sensing image enhancement.
出处
《光学精密工程》
EI
CAS
CSCD
北大核心
2008年第10期2030-2037,共8页
Optics and Precision Engineering
基金
总装备部预先研究基金资助项目(No.404050603)
