基于可见光植被指数的乌海市矿山排土场坡面植被覆盖信息提取研究

李鹏飞; 郭小平; 顾清敏; 张昕; 冯昶栋; 郭光

doi:10.12171/j.1000-1522.20190252

基于可见光植被指数的乌海市矿山排土场坡面植被覆盖信息提取研究

Vegetation coverage information extraction of mine dump slope in Wuhai City of Inner Mongolia based on visible vegetation index

摘要

摘要:
目的利用可见光植被指数快速准确获取矿山排土场坡面植被盖度，为乌海矿区排土场坡面植被调查提供有效方法。
方法选取乌海市典型矿山排土场，通过样方调查法、无人机遥感及可见光植被指数计算筛选适于研究区排土场坡面植被提取的可见光植被指数，并估算其植被盖度，试为排土场坡面植被盖度提取提供新方法。
结果结果表明：（1）不同可见光植被指数提取植被效果存在一定差异，其中绿红比值指数（RGRI）和绿蓝比值指数（BGRI）的灰度图中越暗的部分代表植被指数越大，而其他常见可见光植被指数是越亮的部分代表植被指数越大。（2）研究区中不同可见光植被指数灰度图像特征值基本分布在− 1，1范围内，由蓝、绿波段构建的归一化绿蓝差异指数（NGBDI）和绿蓝比值指数（BGRI）的灰度图中植被与裸地像元值范围有较大重叠，即存在部分混淆。（3）常见可见光植被指数中，可见光波段差异植被指数（VDVI）可以快速准确提取研究区排土场坡面植被，通过人工目视解译及误差矩阵得到VDVI植被指数提取结果平均识别精度在93.4%，表明VDVI植被指数更加适用于乌海市矿山排土场坡面植被提取，优于其他常见可见光植被指数，利用该方法估算可得研究区坡面植被盖度约20.4%。
结论可见光植被指数作为一种非监督分类方法，无需人工选择参考地物即可提取植被，可以作为矿山排土场坡面植被盖度调查的一种新方法，具有广阔的应用前景，同时研究表明VDVI植被指数在提取乌海市矿山排土场坡面植被盖度时具有较高提取精度，对指导当地矿山排土场植被恢复具有实际意义。

Abstract:
ObjectiveThis paper aims to use visible vegetation index to get the slope vegetation coverage of mine dump quickly and accurately, and to provide an effective method for the vegetation investigation of slope in Wuhai mining area, Inner Mongolia of northern China.
MethodThis study selected one of the typical mine dumps in the Wuhai City to choose a suitable visible vegetation index for local research, estimate the dump vegetation coverage, and furthermore try to provide a new method for extracting the dump slope vegetation coverage by means of Quadrat survey, UAV remote sensing and visible vegetation index.
Result(1) Different visible vegetation indexes had varied vegetation extraction effects. Among them, the darker part of red-green ratio index (RGRI) and blue-green ratio index (BGRI) represented the larger vegetation index, while the brighter part of other common visible vegetation indexes represented the larger vegetation index. (2) Gray image feature values of different visible vegetation indexes mainly distributed in range of − 1, 1 in the research area. There was a big overlap of pixel value between vegetation and land in the gray image of normalized green-blue difference index (NGBDI) and BGRI constructed from blue and green bands, which means some partial confusion between them. (3) Visible-band difference vegetation index (VDVI) can extract slope vegetation quickly and accurately among common visible vegetation indices. The average recognition accuracy of VDVI vegetation index was 93.4% by manual visual interpretation and error matrix, showing that VDVI vegetation index could be more suitable for the vegetation extraction on the mine dump in Wuhai City, and was better than the vegetation index of other common visible light. The vegetation coverage of the study area was approximately 20.4% by this method.
ConclusionVisible vegetation index, an unsupervised classified method, could be a new method for investigating vegetation coverage of mine dump slope without selecting reference features by people but extract slope vegetation coverage directly and has a broad application prospect. The study also indicates that VDVI shows higher extraction accuracy of extracting vegetation coverage of mine dump slope in the Wuhai City, which means a practical significance to guide the vegetation restoration in local mine dump.

HTML全文

参考文献(34)

施引文献

资源附件(0)