基于Cycle Spinning Contourlet变换和总变分最小化的遥感图像去噪算法(英文)

Remote Sensing Image Denoising Algorithm Based on Fusion Theory Using Cycle Spinning Contourlet Transform and Total Variation Minimization

针对大部分已有的遥感图像去噪算法在去噪的同时不能有效的保留细节和增强边缘,提出了一种基于Cycle Spinning Contourlet变换和总变分最小化的图像去噪新算法.该算法依据了Cycle Spinning Contourlet变换能够很好的保留原始图像的细节和纹理信息,而总变分最小化方法具有在去噪的同时增强图像边缘的特性,因此使用所提出的融合规则对两种算法去噪后的图像进行融合能够取得更好的增强效果.通过对比,实验结果表明该算法不仅能在很大程度上削弱分别由平移不变Contourlet变换和总变分最小化的图像去噪方法产生的伪吉布斯现象和阶梯效应,而且视觉效果和PSNR值均优于其它方法,同时该算法能够保留更多的光谱信息,因此该算法是一种有效的遥感图像去噪算法.

In order to solve the problem that most of existing image denoising methods insufficiency preserve the details and enhance edges while implementing denoising, a new method for remote sensing image denoising is proposed, based on a combination of cycle spinning contourlet transform （CT）, and the total variation （TV） minimization scheme. The proposed method relies on principles that CT scheme is well suited for preserving detailed and fine textures information of original image while TV minimization denoising scheme is capable of enhancing sharpened significant edges while denoising, therefore to fuse the two schemes using the proposed fusion rule can achieve better results. Compared with several commonly used approaches, the experimental results show that this novel algorithm is capable of reducing Gibbs phenomenon and staircase effect produced by CT and TV denoising methods respectively, superior both in visual quality of denoising and Peak Signal to Noise Ratio （PSNR）, and preserves more spectral information and less spectral distortion simultaneously.