SPOT 5在森林资源二类调查应用中数据优化研究  

A Study on the Data Optimization of SPOT 5 in Secondary Level Survey of Forest Resource

作　　者：何志明[1,2] 李月臣[1,2] 韩建文[3] 胡晓明[1,2] 杨清钰[3] 

机构地区：[1]重庆师范大学地理与旅游学院,重庆400047 [2]GIS应用研究重庆市高校重点实验室,重庆400047 [3]重庆市林业科学研究院,重庆400036

出　　处：《遥感信息》2012年第4期57-63,70,共8页Remote Sensing Information

基　　金：重庆市博士后科研项目特别资助基金(No渝xm201102001);重庆市气象局开放基金(Nokfjj-201103);资源环境与生态建设重庆市高校创新团队基金

摘　　要：为了实现SPOT 5影像应用于森林资源二类调查过程中影像数据最优的目视效果,首先利用标准差、均值、信息熵、OIF指数等指标对影像融合及目视解译最佳波段进行选取;其次,针对性地引入了亮度信息、空间细节和纹理结构、信息量、光谱特性保持程度等融合效果评价参数,并以高空间分辨率信息继承程度作为重要参数以反映融合影像空间分辨率特征,把最佳波段作为统计对象对常用的9种像素级融合方法进行了融合效果分析。比较得出PCA融合为9种方法中综合最优,影像亮度、空间分辨率信息较高,视觉效果最佳;乘积融合和Brovey融合虽然空间分辨率信息继承程度等指标也较优,但是光谱扭曲太大。小波融合等方法总体光谱特征普遍较好,但是噪声较多,空间分辨率信息不佳,不适用于目视解译。定量评价之后的定性评价分析进一步强调了高空间分辨率信息继承程度做为评价指标的重要性。In order to achieve the optimal visual effect in secondary level survey of forest resource using SPOT 5 Image,it should be considered two aspects.Firstly,standard deviation,mean value,information entropy and OIF index are used to indicate the best bands in image fusion and visual interpretation.Secondly,five parameters are introduced such as the brightness,spatial details and texture structure,amount of information,spectral characteristics and more significant parameters which reflect the spatial resolution features of fused images.Three optimal bands are used to estimate the fusion effects of 9 pixel-based methods.The conclusion shows that PCA fusion is the optimal fusion method among 9 kinds of methods,its image brightness,spectral details and spatial resolution are excellent.Although the Multiplicative and Brovey methods are clearer too,but the extent of spectral changes with the original multi-spectral images varied a lot.The spectral characteristics of other methods are also excellent,but more noise,low spatial resolution,what make them unsuitable for visual interpretation.

关 键 词：SPOT 5 遥感影像 波段选择 融合 森林资源二类调查 数据优化 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置]