基于最优尺度的生产建设扰动图斑识别

发布时间：2018-05-04 13:20

本文选题：生产建设项目 + 扰动图斑　；参考：《中国水土保持科学》2017年06期

【摘要】：遥感识别生产建设扰动图斑对水土保持监测具有重要的意义。本文尝试提出一套基于高分辨率影像自动获取生产建设扰动图斑的方法,重点探讨扰动图斑最优分割尺度的自动获取方法,以解决当前过度依赖分类者的多次尝试和主观判断问题。以高分1号影像为例,利用局部方差方法和目标函数方法计算生产建设扰动图斑的最优分割尺度,结果显示,与目视判读的最优分割尺度对比,目标函数方法的结果精确,局部方差方法的结果略微偏大。在此基础上,利用面向对象的监督分类方法,实现了对扰动图斑的识别。利用现场调查的结果验证,本文中的方法较为可靠,制图精度和用户精度分别达到86.3%和84.2%,高于其他分割尺度的识别精度。
[Abstract]:It is very important for soil and water conservation to identify the disturbance map of production and construction by remote sensing. This paper attempts to propose a method for automatic acquisition of production and construction disturbance spots based on high-resolution images, with emphasis on the automatic acquisition method of optimal segmentation scale of disturbance patterns. In order to solve the current excessive reliance on the classification of multiple attempts and subjective judgment. Taking Gaofen 1 image as an example, the local variance method and objective function method are used to calculate the optimal segmentation scale of production disturbance map. The results show that compared with the optimal segmentation scale of visual interpretation, the result of objective function method is accurate. The results of the local variance method are slightly larger. On this basis, the object-oriented supervised classification method is used to realize the recognition of disturbance patterns. The results of field investigation show that the method in this paper is reliable. The accuracy of cartography and user is 86.3% and 84.2% respectively, which is higher than the recognition accuracy of other segmentation scales.
【作者单位】：珠江水利委员会珠江水利科学研究院;珠江水利委员会珠江流域水土保持监测中心站;水利部水土保持监测中心;
【基金】：水利部科技推广项目“水土保持监督监测现场定量信息采集移动平台推广应用”(SF-201606);水利部综合事业局拔尖人才专项资金项目“生产建设项目水土保持天地一体化监控技术研究”(20150608-01);水利部预算项目“岩溶地区石漠化现状及变化趋势”(1261520610102);水利部预算项目“全国水土流失动态监测与公告”(1261520610101)
【分类号】：S157

【相似文献】