高级检索

    基于K-T变换的地表水体信息遥感自动提取模型

    Automatic water bodies extraction model based on K-T transformation

    • 摘要: 传统的水体提取方法阈值法和谱间关系法对于水体和云体阴影难以区别,影响提取精度。为建立一种能精确提取水体并且能够消除各种云体阴影影响的模型,对所研究的图像先进行K-T变换,再利用新生成的6个波段建立水体信息提取模型。结果表明:只有水体(Water)的波谱曲线具有W型的独特特征,建立的水体提取模型Greenness〈Wetness〉Fourth不但可消除云体阴影,而且可高精度地提取地表水体。

       

      Abstract: Threshold method and spectral relationship method are traditional methods for water body information extraction, the two methods, however, could not distinguish cloud shadow from water body, which leads to lower precision. For achieving a optimal model, the K-T transformation(Kauth-Thomas transformation)was applied to the TM image firstly, and then the improved model of water bodies extraction was found using the six new wave bands of the TM image after K-T transformation. It can be concluded that only when the spectral curve of water body occurred with W shape, could the new model of water body information extraction, Greenness 〈Wetness〉 Fourth, eliminate the influence of all shadow and extract the water body with higher precision.

       

    /

    返回文章
    返回