论文部分内容阅读
基于非局部自相似性的遥感图像稀疏去噪方法研究,在为后续的图像分析、识别以及较高层次的处理提供保证方面具有重要意义。针对遥感图像中存在非局部自相似性和稀疏性,在分析传统稀疏去噪模型的基础上,将具有相似结构的非局部块构建成组,用组作为稀疏表示单元,利用基于组正则化稀疏模型进行图像去噪。此外,针对采用整幅图像进行字典学习具有高计算复杂度,分析组特点,为每个组自适应学习一个字典。最后,为获得有效的去噪结果,利用迭代收缩阈值算法解决L0最小化问题。以“资源三号”遥感图像为数据进行实验,结果表明,该算法能较好地去除遥感图像的噪声,提高图像的峰值信噪比,保持图像结构信息。基于非局部自相似性的遥感图像稀疏去噪算法能够充分利用图像块信息有效的去除图像中的噪声,提高图像质量。
The study of sparse denoising based on non-local self-similarity in remote sensing imagery is of great significance in providing assurance for subsequent image analysis, recognition and higher-level processing. Aiming at the non-local self-similarity and sparsity in remote sensing images, based on the analysis of traditional sparse denoising models, the nonlocal blocks with similar structure are built into groups and the groups are used as sparse representation units. Model for image denoising. In addition, dictionary learning using the entire image has high computational complexity, analyzing group characteristics and learning a dictionary adaptively for each group. Finally, in order to obtain an effective denoising result, an iterative shrinkage threshold algorithm is used to solve the L0 minimization problem. The experimental results show that this algorithm can remove the noise of remote sensing image, improve the peak signal to noise ratio of the image and maintain the image structure information. Remote sensing image sparse denoising algorithm based on non-local self-similarity can make full use of image block information to effectively remove the noise in the image and improve the image quality.