期刊文献+

基于小波融合的泉州市QuickBird影像绿地信息提取 被引量:1

Greenbelt Information Extraction of QuickBird Image Based on Wavelet Fusion in Quanzhou City
在线阅读 下载PDF
导出
摘要 以2006年泉州市部分QuickBird多光谱影像与全色影像为实验区数据基础,在遥感软件ERDAS IMAGINE 9.2平台支持下,应用小波融合算法及融合规则实现QuickBird融合处理,在此基础上进行归一化植被指数(NDVI)运算,通过人机交互式提取得到实验区5种主要的绿地类型:公园绿地、生产绿地、防护绿地、附属绿地及其他绿地.结果表明:1)QuickBird多光谱影像、全色影像的信息熵分别为3.975,4.162,而QuickBird融合影像的信息熵为7.251,明显高于前两者,融合后空间信息更丰富;2)QuickBird融合影像的平均梯度是3.328,多光谱影像、全色影像分别是1.552,2.965,融合影像纹理特征更明显;3)QuickBird融合影像与多光谱影像的偏差是0.035 9,全色影像的偏差为0.0562,均接近于0,小波融合影像较好地保持了源图像的光谱特性;4)绿地信息分类提取的精度为89.6%,趋近90%. On the experimental data basis of QuickBird multi-spectral image and panchromatic image of Quanzhou City in 2006, QuickBird fusion processing was realized by using wavelet fusion algorithm and fusion rule on the support of ERDAS IMAGINE 9.2. And normalized differential vegetation index (NDVI) was operated to extract five greenbelt types : park greenbelt, productive greenbelt, protective greenbelt, attached greenbelt and other greenbelts under the method of human-computer interaction. The results showed that: 1 ) The entropy of QuickBird multi-spectral image was 3. 975, and the entropy of QuickBird panchromatic image was 4. 162. While the entropy of QuickBird fusion image was 7. 251, it was higher than two former. Spatial information was more abundant after fusion. 2) The average gradient of QuickBird fusion image was 3. 328. But the average gradient of multi-spectral image was 1. 552, and the average gradient of panchromatic image was 2. 965. Texture feature of fusion image was more obvious. 3 )The deviation between QuickBird fusion image and multi-spectral image was 0. 035 9, and the deviation between QuickBird fusion image and panchromatic image was 0. 056 2. They were close to zero. So wavelet fusion image maintained well with spectral properties of source images. 4) The precision of greenbelt information extraction was 89.6%, it was close to 90%.
出处 《北华大学学报(自然科学版)》 CAS 2013年第6期737-743,共7页 Journal of Beihua University(Natural Science)
基金 福建省科技厅重点资助项目(2009N0009) 国家自然科学基金青年项目(41301203) 国家自然科学基金面上项目(31370624 30870435) 福建省自然科学基金项目(2011J01071 2008J0116)
关键词 QUICKBIRD影像 小波融合 NDVI 绿地信息 泉州 QuickBird image wavelet fusion NDVI greenbelt information Quanzhou
作者简介 郑晓燕(1988-),女,硕士研究生,主要从事景观生态学与地理信息系统研究; 通信作者:何东进(1969-),男,教授,博士,博士生导师,主要从事景观生态学与森林生态学研究.
  • 相关文献

参考文献14

二级参考文献174

共引文献130

同被引文献4

引证文献1

二级引证文献3

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部