摘要
Brushlet是一种新的图像方向信息分析工具,已被成功应用于图像融合与纹理分类等领域。提出一种基于Brushlet变换的多层阈值选择策略,并将其应用到SAR图像相干斑去噪中,通过对真实SAR图像的去噪实验表明,相比于传统的Wavelet方法,Brushlet变换域的多层阈值算法能获得更好的去噪效果,同时较好地保留了图像细节特征并获得更好的等效视数(ENL)。
Brushlet is a novel tool for image directionality analysis,which is adopted in image fusion and texture classification. A SAR image denoising algorithm based on multi - level threshold selection using Brushier is proposed. The denoising experimentson real SAR image prove that the proposed method outperforms the conventional Wavelet methods. The SAR image details can be well preserved and better ENL can be achieved by multi - level threshold via Brushlet.
出处
《现代电子技术》
2009年第8期97-99,102,共4页
Modern Electronics Technique
基金
国家自然科学基金资助项目(50675079)
