基于面向对象的西双版纳橡胶林提取

面向对象遥感信息提取技术可以综合光谱、空间、纹理等特征,从而提高影像分类的精度。以西双版纳地区为研究对象,利用相关数据,以面向对象方法为核心,通过图像预处理、图像增强、光谱曲线提取与特征提取、分割、图像分类等技术环节,对其土地利用信息进行提取研究,尤其是橡胶林的提取。探索了该地区利用遥感数据为信息源快速获取土地利用的有效方法,并用2期数据研究区内的土地利用变化情况。最后,得出面向对象的方法在该研究区域是可行的,而且也相对提高了分类精度,同时,提取了土地利用变化情况,为该地区综合研究提供基础信息,而且更重要的是从另一个侧面说明了面向对象分类的优越性。

Extraction of Rubber Forest in Xishuangbanna Based on Object-oriented Method

Object-oriented remote-sensing information extraction technology can integrate spectrum,space,texture and oth-er characteristics to improve the accuracy of image classification. Using the relevant data,taking object-oriented method as the core,the author studied the extraction of land use information（ especially rubber forest information）in Xishuangbanna area through such technological links as image preprocessing,image enhancement,spectral curve extraction,feature extraction,segmentation and image classification. This study explored the effective method for the rapid acquisition of land use information in this area according to remote-sensing data,and analyzed the land use changes in the study area based on 2-term data. Finally,it was concluded that the object-oriented method was feasible and superior in image classification in this area,and it could relatively improve the classifi-cation precision. At the same time,the change situation of land use in Xishuangbanna area was extracted with this method,which can provide basic information for the comprehensive study of this area.