Landsat_8多光谱波段纹理对叶面积指数的影响分析

Impact of Multispectral Bands Texture on Leaf Area Index Using Landsat_8

以北京密云县为例,分析植被纹理特征与叶面积指数的相关性,为叶面积指数模型引入纹理特征提供参考。利用Landsat_8影像多光谱波段间的相关性,将相关性最小的波段4、波段5分别与全色波段融合后提取纹理特征,研究融合后纹理特征对叶面积指数的影响。结果表明,融合后纹理特征与叶面积指数的相关性多数显著高于融合前,其中针叶林融合后纹理特征与叶面积指数的相关系数比阔叶林的高。但在研究阔叶林波段5融合后的纹理值对叶面积指数的影响中,其相关性却比融合前的低。说明融合后的纹理值虽然对叶面积指数模型的构建具有重要作用,但也需考虑不同的林分、波段对叶面积指数的影响。

The correlation between forest texture features and leaf area index was analyzed and discussed to provide a reference for the introduction of texture features into leaf area index model in Miyun County of Beijing.Through correlation analysis among multiple spectral bands of Landsat_8 image,the band 4 and band 5 which had a minimum correlation were combined with panchromatic band to extract texture features.Then the influence of texture features after fusion on leaf area index was discussed.The results showed that the correlation of texture features after fusion and leaf area index was significantly higher than that of texture features before fusion and leaf area index.Furthermore,the correlation coefficient of coniferous forest after fusion was larger than that of broadleaf forest after fusion.However,with the research that texture feature of band 5 after fusion had an effect on leaf area index in broadleaf forest,their correlation was lower than that of before fusion.It was illustrated that texture value after fusion played an important role in the construction of leaf area index model on the whole,but different stand and bands were also considered in the influence of leaf area index.