纹理特征是遥感影像的重要特征之一。本研究以湖南省桃江县为研究对象,利用CBERS-02B星为遥感数据源,采用灰度共生矩阵法提取影像纹理特征。根据纹理参数的标准差和各参数间的相关性,对纹理参数进行了筛选,并对竹林和针叶林的均值和方差的季相变化特征进行了统计和分析。结果表明:在5月、8月和12月之间,竹林均值逐渐减小而针叶林在8月份出现峰值,针叶林标准差逐渐增大,而竹林在8月份出现峰值。
Texture feature is a significant factor in remote sensing imagery. The aim of this analysis was extract texture feature variations of bamboo and coniferous forests based on CBERS-02B Data in Taojiang County in Hunan Province with gray level co-occurrence matrix. The texture parameters were selected by using the standard deviation and correlation methods. The laws of seasonal variation of the mean parameter and variance parameters were studied in detail by using statistical analysis. The analysis showed that the average of the mean parameter of bamboo forests gradually decreased and the standard deviation of variance parameter of coniferous forests gradually increased in May, August and December. The maximum average of the mean parameter of coniferous forests and the maximum standard deviation of variance parameter of bamboos occur in August.