应用空间统计学理论解译遥感影像信息“缺失”区

Interpretation of Pixel-missing Patch of Remote Sensing Image with Kriging Interpolation of Spatial Statistics

借助图像处理软件ERDAS、地理信息系统软件ArcInfo以及空间统计分析软件ILWIS ,在对TM遥感影像进行分类的基础上 ,运用空间统计学理论以及Kriging插值技术 ,内插了影像真实信息“缺失”斑块的信息 ,插值结果通过了精度检验。为解译影像信息“缺失”区 ,提供了一种手段和方法。

The purpose of this paper was mainly to interpret the pixel missing patch of image by using the kriging interpolation technology of spatial statistics. The TM image (a resolution of 30m×30m) of the Jingouling Forest Farm of the Wangqing Forestry Bureau of Northeast China on 1 July, 1997 was selected as data resource, and some patches with cloud and cloud shade from the TM image were adopted as experimental data in this paper. Based on the classification for the TM remote sensing image, the pixel information of real pixel missing patches of image (the real pixel missing patch is the patch with the unknown real information of partial regions on the image, which is caused by the shelters of cloud, cloud shade, soot and others) was restored by the kriging interpolation technology of spatial statistics theory under the image processing software—ERDAS, the geographic information system software—ArcInfo and the spatial statistics analysis software—ILWIS. The restored pixel information was reclassified, and then compared with the results of forest resource survey, thus the results (i.e. the classified precision of the restored information of each patch was more than 75%), was obtained. Moreover, the kriging interpolation technology used in this paper grasped the fact that the objects on the global surface possessed continuity and self correlation, and therefore it had better estimation effects. Additionally, the applicable scopes and the applicable conditions of the kriging interpolation were discussed in detail from different aspects. In summary, this paper would provide a method and means for interpreting the pixel missing patch of image.