文献类型：期刊文献

中文题名：GF-6卫星WFV数据在林地类型监测中的应用潜力

英文题名：Potential Application of GF-6 WFV Data in Forest Types Monitoring

作者：刘晋阳[1] 辛存林[1] 武红敢[2] 曾庆伟[3] 史京京[4]

第一作者：刘晋阳

机构：[1]西北师范大学地理与环境科学学院;[2]中国林业科学研究院资源信息研究所;[3]二十一世纪空间技术应用股份有限公司;[4]国家林业局调查规划设计院

年份：2019

卷号：40

期号：2

起止页码：107-116

中文期刊名：航天返回与遥感

外文期刊名：Spacecraft Recovery & Remote Sensing

收录：CSTPCD;;CSCD:【CSCD2019_2020】；

基金：国家自然科学基金(41262001);甘肃省科技支撑基金(1104FKCA116)

语种：中文

中文关键词：宽视场;“高分六号”卫星;随机森林;林地类型;应用潜力;卫星遥感应用

外文关键词：wide field of view (WFV);GF-6 satellite;random forest algorithm;forest types;potential application;satellite remote sensing application

分类号：TP79

摘要：为了了解"高分六号"(GF-6)卫星宽视场(WFV)传感器1A级产品数据品质,评估新增4个波段在林地类型分类研究中的应用潜力,文章以内蒙古鄂伦春自治旗的秋季GF-6卫星WFV数据为对象,开展了绝对定位精度的测试,并应用随机森林方法进行了林地类型分类研究。结果表明,GF-6卫星的WFV平原区数据几何畸变较小,无控制点有理多项式系数校正后图像配准的中误差在1个像元之内,有助于降低遥感技术应用的门槛,提高工作效率。新增的红边1、红边2、黄边波段对林地类型分类影响显著,整体分类精度从74%提高到80%以上,紫边波段对于森林类型分类也有所贡献。因此,GF-6卫星WFV的8波段多光谱遥感数据对中国大尺度森林资源、森林灾害等的宏观动态监测具有重要意义和极大应用潜力。
In order to understand the data quality of the 1A grade product of WFV (Wide Field of View) sensor on GF-6, we evaluated the potential application of the new four bands in forest types classification. Using the autumn GF-6 satellite data of the Inner Mongolia Oroqen Autonomous Banner. As the object, the absolute positioning accuracy test is carried out, and the forest type classification is studied using the random forest method. The results show that the geometric distortion of the GF-6 WFV data is small in the plain area, and the error of image registration after uncontrolled RPC correction is within 1 pixel, which helps to reduce the threshold of remote sensing technology application and improve the working efficiency. The newly added red edge 1, red edge 2 and yellow edge bands have significant influence on forest type classification, and the overall classification accuracy has increased from 74% to over 80%. The purple edge band also contributes to forest type classification. Therefore, the 8-band multi-spectral remote sensing data of the GF-6 WFV is of great significance and great potential for macro-dynamic monitoring of large-scale forest resources and forest disasters in China.