基于支持向量机的遥感图像厚云去除算法  

The Algorithm for Removing Thick Clouds in Remote Sensing Image based on Support Vector Machine

作　　者：唐王琴[1,2] 梁栋[1,2] 胡根生[1,2] 马雪亮[1,2] 杭丹萍[1,2] 

机构地区：[1]安徽大学计算智能与信号处理教育部重点实验室,安徽合肥230039 [2]安徽大学电子信息工程学院,安徽合肥230039

出　　处：《遥感技术与应用》2011年第1期111-116,共6页Remote Sensing Technology and Application

基　　金：国家自然科学基金(60772121);安徽省教育厅重点科研计划资助项目(KJ2010A021);安徽大学"211工程"学术创新团队基金资助

摘　　要：厚云的存在大大降低了遥感图像的利用率,利用支持向量机超强捕获边缘点的能力和图像融合方法,提出了一种基于支持向量机遥感图像厚云去除算法。首先构造支持向量值轮廓波变换并对图像进行分解,然后进行云层检测和图像融合,最后进行支持向量值轮廓波逆变换,得到重构图像。仿真实验表明,对于有厚云覆盖但无云区重叠的遥感图像,该算法能取得满意的去云效果,不仅保留了图像边缘信息,而且有效地解决了云层残留问题。Thick cloud in remote sensing image greatly reduce the utilization of the images.Using the ability of support vector machine to capture the edge points and image fusion scheme,in this paper,the algorithm based on support vector machine is presented to remove thick clouds in remote sensing image.First of all,support vector value contourlet transform is constructed and the images are decomposed.Then cloud detection and image fusion are carried out.Finally the reconstructed images are got by support vector value contourlet inverse transform.The experimental results reveal that the algorithm can achieve satisfactory result for the remote sensing images with the thick clouds but no overlapping in cloud.Not only the edge information of image is preserved but also the problem of residual clouds is resolved effectively.

关 键 词：支持向量机 云层检测 厚云去除 图像融合 

分 类 号：TP75[自动化与计算机技术—检测技术与自动化装置]