【摘要】 针对高分辨率可见光波段影像植被提取时,因缺少近红外波段引起参考不足的问题,利用植被在可见光方面的光谱特性,实现对可见光波段植被自动化提取与初步分类。在可见光植被指数已有研究的基础上,利用第三次全国土地调查的0.5m正射高分辨率DOM,将不同植被指数带入计算,从一系列可见光植被指数中选取一组最优植被指数。通过筛选,选用RGBVI指数进行实验。检测结果证明,RGBVI-Ostu方法在第三次全国土地调查底图中提取林地植被精度高达91.26%,由此证明该方法使高分辨率可见光植被指数林地识别更精确。更多还原

【Abstract】 In the process of vegetation extraction from high-resolution visible band images,the lack of near-infrared band leads to the problem of insufficient parameters. The automatic extraction and preliminary classification of visible band vegetation are realized by using the spectral characteristics of vegetation in visible light. Based on the existing research of visible light vegetation index,we used the0.5 m ortho high resolution DOM of the third nationl land survey. Different vegetation indexes are substituted into the calculation,and a group of optimal vegetation indexes are selected from a series of visible light vegetation indexes. Through screening,the index of RGBVI was selected for the experiment. The results show that the accuracy of this method is 91.26% in the third nationl land survey map,which proves that the method can make the forest identification more accurate with high-resolution visible light.更多还原