1种遥感影像数据融合的差异性K-TH阶中心矩方法

A Differential K-TH Central Moment Method for Remote Sensing Image Data Fusion

探讨基于K-TH阶中心矩重建一系列基于多波段向量特征和均值向量差异图像的差异性K-TH阶中心矩中心方法。与多波段灰度图像、RGB融合图像、eCongnition软件处理后图像相比结果表明,重建图像信息量在低k值下得到显著增强,可应用于农业病虫害监测、森林火灾、环境污染、洪水治理方面,用于检测与周边环境变异较大区域;当k值增加时,重建图像信息量急剧降低,多波段向量特征和均值向量差异被放大,图像趋向于二值图像,可用于遥感分割、分类、微小差异性检测。

The differential K-TH central moment method was constructed based on multiband vector and mean vector.Compared with the multiband grey images,RGB fusion image and the images processed by eCognition software,the information of the reconstructed images significantly enhanced under low k value,and it could be applied to the monitoring of agricultural insects pest and plant diseases and the control of forest fire,environmental pollution and flood treatment in those areas which differed largely with the surrounding areas;with the increasing k value,the information of reconstructed images reduced substantially,the difference between multiband vector and mean vector was enlarged,the images tended to be binary images and was applicable to RS segmentation,classification and detection of micro difference.