|
|||||
|
|
| 城市景观遥感影像融合质量对小波基选取的响应 | |
| 引用本文： | 陈颖彪,郭冠华,吴志峰,魏建兵.城市景观遥感影像融合质量对小波基选取的响应[J].地理与地理信息科学,2011,27(4):98-102,113. |
| 作者姓名： | 陈颖彪 郭冠华 吴志峰 魏建兵 |
| 作者单位： | 1. 广州大学地理科学学院,广东广州510006; 2. 广州大学地理科学学院,广东广州510006;广东省生态环境与土壤研究所,广东广州510650 3. 广东省生态环境与土壤研究所,广东广州,510650 |
| 基金项目： | 中国科学院资源与环境信息系统国家重点实验室开放研究基金，国家自然科学基金，广东省自然科学基金 |
| 摘 要： | 应用IHS和小波变换结合的融合算法时,小波基的选取是影响融合图像质量的关键,而且包含不同地物信息的影像融合质量对小波基的响应特征有待深入探讨。该文以SPOT全色和多光谱影像为数据源,选取大量高、低密度建筑城市景观影像样本,用信息量、平均梯度和偏差指数3个指标定量评价融合质量,分析不同小波簇下两类融合图像之间的质量差异。研究表明:融合质量与影像所含地物特征密切相关;相对于其它小波簇,rbioNr.Nd表现出较强的变异性;两类融合图像在各小波簇上的波动性差异明显,并呈分段特征;根据质量需求和具体景观特征影像,可以在分解层数固定的情况下选取最佳小波基,以获取高质量融合图像。 |
| 关 键 词： | 影像融合 高、低密度建筑城市景观 小波变换 小波基 |
Response of Urban Landscape Images Fusion to the Wavelet Basis Selection |
|
| CHEN Ying-biao,GUO Guan-hua,WU Zhi-feng,WEI Jian-bing.Response of Urban Landscape Images Fusion to the Wavelet Basis Selection[J].Geography and Geo-Information Science,2011,27(4):98-102,113. | |
| Authors: | CHEN Ying-biao GUO Guan-hua WU Zhi-feng WEI Jian-bing |
| Institution: | 1.School of Geographical Sciences,Guangzhou University,Guangzhou 510006;2.Guangdong Institute of Eco-environment and Soil Sciences,Guangzhou 510650,China) |
| Abstract: | Wavelet basis selection plays an important role in the fusion algorithm of integration of wavelet transform and IHS in remote images fusion.However,response of remote images which contain various land use or landscape information to the wavelet basis selection should obtain more attention when this fusion method is used.In this paper,to deeply explore the difference performance of fusion results from different landscape images by diversified clusters of wavelet and wavelet basis,lots of high and low density buildings of urban landscape images samples were cut out from SPOT panchromatic data and multi-spectral data.The integration of wavelet transform and IHS was adopted and several indices for images performance including entropy,average gradient and deviation were calculated.Results showed that the fusion images performance have a close relationship to the landscape information included in the raw images,and the performance showed quite different according to indices.The variability of performance with rbioNr.Nd using showed more remarkable than the others clusters of wavelet to these two kinds of landscape images.The fluctuate characteristic of indices curves for tow types landscape images showed obvious difference.And those differences just arise in the given series of wavelet basis.With the analysis giving from this paper,the best wavelet basis could be chose when the specific landscape images and composition level was set.This paper supports better understanding of the fusion algorithm and makes its application simpler. |
| Keywords: | images fusion high/low density buildings urban landscape wavelet transform selection of wavelet basis |
| 本文献已被 CNKI 万方数据 等数据库收录！ | |