CIVCO地形校正模型的改进及其应用

在地形复杂的山区,地形效应造成了地物的光谱特征扭曲,严重地影响了地物的识别与分类。地形校正不仅能提高遥感识别与分类的精度,还是各种定量化遥感应用的前提。CIVCO模型是一种基于统计的经验地形校正模型,原模型中阴坡和阳坡需要手工选取,且校正效果随坡度而异。通过利用太阳方位角和坡向简化了阴坡和阳坡的选取,并按照坡度分级,分别求出校正系数。在模型改进的基础上,选择黑河上游祁连山区的TM影像和相应的DEM数据进行试验,结果表明,与C模型及原模型相比,改进的模型取得了更好的校正效果。

Improved CIVCO Topographic Correction Model and Application

The object spectral is seriously disturbed by the topographic effects in mountainous area, which severely affected the object recognition and classification. Topographic correction not only improves the classification accuracy but is a prerequisite in any quantitative remote sensing with high preciseness approach. CIVCO model is an experiential and applied model, based on the statistics. It is complex and arduous to select sunlit and shade, and the correct result is affected by slope. In this paper, we use sun azimuth and aspect to simplify the selection of sunlit and shade, grade the gradient and compute the correction coefficient. Based on the improved model, the experiment using a Landsat TM image of the Qilian Mountains in the upriver area of Heihe and the corresponding DEM data proved that the improved method get better results than the C and CIVCO model.