基于低空可见光谱的植被覆盖率计算

目的]将低空遥感技术应用到水土保持设施验收中,从可见光谱遥感影像提取区域植被信息,提出可准确、客观计算水土保持设施验收指标中植被覆盖率的方法,以减少工作人员的业外工作量,提高测算效率。方法]将仅含有可见波谱信息的低空遥感图像作为研究对象,在利用植被指数红绿比指数(RGRI),过绿指数(EXG),可见光波段差异植被指数(VDVI),归一化绿蓝差异指数(NGBDI)和归一化绿红差异指数(NRGRDI)分析图像波谱特性的基础上,采用双峰直方图法和最大熵值法确定各植被指数的阈值,再使用ENVI软件提取图像的植被信息,并计算植被覆盖率,与参照结果进行比对。结果]利用可见光谱差异指数(VDVI)提取的植被信息精度高达95.32%。由此计算得到的植被覆盖率为54.43%,与实际情况最为接近。结论]基于可见光谱遥感影像计算植被覆盖率的方法具有可行性,该方法人工干预少,结果准确度高,可为水土保持设施验收提供实时的数据支撑。

Vegetation Coverage Calculation Based on Low Altitude Visible Spectrum

Objective] Applying low altitude remote sensing to the acceptance of soil and water conservation facilities so as to extract vegetation information based on visible spectrum from remote sensing image, and to propose an accurate and objective method to calculate the coverage rate which is an indicator of the soil and water conservation facilities evaluation in the hope of reducing workload and improving effectiveness.Methods] The spectral characteristics of low-altitude remote sensing images containing only visible spectral information were analyzed by five vegetation indices as RGRI (ration vegetation index), EXG (excess green),VDVI (visible-band difference vegetation index),NGBDI (normalized green-blue difference index) and NRGRDI (normalized green-red difference index). And the threshold of each vegetation index was determined by the maximum entropy method or bimodal histogram method. Furthermore, with the help of ENVI, the vegetation information was extracted and the vegetation coverage were then calculated and compared with references.Results] The accuracy of vegetation information extracted from the visible-band difference vegetation index (VDVI) was as high as 95.32%, and the vegetation coverage was 54.43%, which was the closest to the actual value.Conclusion] It is feasible to calculate vegetation coverage from remote sensing image based on visible band. The method can provide real-time data as supporting information for the acceptance assessment of soil and water conservation facilities with its advantage of few artificial intervention and high accuracy.