| 李飞,卢湖川,薄纯娟.基于上下文多字典学习的高光谱波段选择[J].,2021,61(1):104-110 |
| 基于上下文多字典学习的高光谱波段选择 |
| Hyperspectral band selection based on context multi dictionary learning |
| |
| DOI：10.7511/dllgxb202101014 |
| 中文关键词: 高光谱图像 波段选择 稀疏表示 |
| 英文关键词: hyperspectral images band selection sparse representation |
| 基金项目:国家自然科学基金联合基金资助项目(U1903215). |
|
| 摘要点击次数: 238 |
| 全文下载次数: 388 |
| 中文摘要: |
| 高光谱图像存在大量冗余信息，波段选择是一种有效的减少冗余、降低光谱维数的方法．提出一种基于上下文多字典学习的高光谱图像波段选择算法．该算法使每个波段的图像都可以通过其他波段图像的线性组合来近似表示，而且能够保证相邻波段图像具有相似的性质．同时通过稀疏求解方法求出每个波段对应整个原始图像的权重，便可按照权重来选择波段．实验结果表明，该算法相对其他波段选择算法具有良好的技术性能． |
| 英文摘要: |
| There are a lot of redundant information in hyperspectral images. Band selection is an effective method to reduce redundancy and reduce spectral dimension. A new band selection algorithm for hyperspectral image based on context multi dictionary learning is proposed. Under this new algorithm, each band image can be approximately represented by a linear combination of other band images, and adjacent band images have similar properties. Using the sparse solution method, the weight of each band corresponding to the whole original image is obtained, and bands can be selected by their weights. Experimental results show that the proposed algorithm has better performance than other band selection algorithms. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
|
|
|