Knowledge Commons of Institute of Automation,CAS
HYPERSPECTRAL DATA RECOVERY WITH THE GRADIENT FIELD OF COINCIDENT PANCHROMATIC IMAGERY | |
Ma L(马雷)1,2![]() ![]() | |
2014 | |
会议名称 | 2014 IEEE International Geoscience and Remote Sensing Symposium |
页码 | 2946-2949 |
会议日期 | 2014-07-13 |
会议地点 | Québec, Canada |
摘要 | Hyperspectral Images often suffer from missing pixels due to acquisition system problem. Missing pixels in images are usually tackled by means of interpolation methods by using neighborhood known pixels, but once the missing region is large, there may not be sufficient information in the neighborhood to reconstruct well. Fortunately, hyperspectral sensors are also often flown with boresighted, higher resolution panchromatic sensors. We propose a novel hyperspectral data recovery algorithm based on multi-source image fusion, guided by the gradient field of coincident panchromatic image in inpainting processing. The experiment results are illustrated with practical examples and verify the efficacy of this algorithm. |
关键词 | Hyperspectral Data Image Recovery Poisson Inpainting |
收录类别 | EI |
语种 | 英语 |
文献类型 | 会议论文 |
条目标识符 | http://ir.ia.ac.cn/handle/173211/40668 |
专题 | 复杂系统认知与决策实验室_高效智能计算与学习 |
作者单位 | 1.Institute of Automation, Chinese Academy of Sciences 2.Beijing Institute of Remote Sensing Information, Beijing |
第一作者单位 | 中国科学院自动化研究所 |
推荐引用方式 GB/T 7714 | Ma L,Bitao Jiang,Xinwei Jiang. HYPERSPECTRAL DATA RECOVERY WITH THE GRADIENT FIELD OF COINCIDENT PANCHROMATIC IMAGERY[C],2014:2946-2949. |
条目包含的文件 | ||||||
文件名称/大小 | 文献类型 | 版本类型 | 开放类型 | 使用许可 | ||
Hyperspectral Data R(7658KB) | 会议论文 | 开放获取 | CC BY-NC-SA | 浏览 |
除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。
修改评论