摘要
为了探讨与分析无人机大样方估算草地植被覆盖度的可行性及植被覆盖度与生物量的相关性,利用大疆Phantom3Professiona无人机获取呼伦贝尔草甸草原60 m×60 m大样方数据,利用开关中值滤波及基于维纳滤波的小波变换对图像进行去噪处理,并用直方图均衡化进行图像增强,构建过绿指数与颜色指数,根据最大熵原理,建立植被覆盖度与背景的双阈值目标函数,采用遗传算法寻找最佳分割阈值,从而估算草地植被覆盖度。结合野外同步实地照相法获取1 m×1 m样方数据,利用t检验及计算均方根误差验证基于无人机大样方估算草地植被覆盖度的精度。结果表明,利用过绿植被指数基于遗传算法的最大熵法估算无人机大样方草地植被覆盖度的精度较高(t检验P值为0.272,均方根误差为6.635 6);株高及覆盖度的交叉变量构建的拟合方程效果(R2为0.853 6,RMSE为2.442 0g·m-2,估算精度为83.41%)优于单一变量拟合方程。利用无人机数据及最大熵-遗传算法不仅可以快速、准确、高效地获取植被地面信息,同时也可以为遥感估算植被覆盖度提供地面验证,从而提高植被覆盖度估算模型的精度及模型验证的准确性。
To investigate the potential ability of unmanned aerial vehicle(UAV)in the estimation of grassland coverage and its relation to the biomass,in this study,the UAV of Phantom 3 Professiona was used to acquire major quadrat with 60 m×60 m from Hulunbuir pasture land.The Wiener filtering and wavelet transformation methods were applied to de-noise UAV imagery.The imagery was further enhanced by histogram equalization.Furthermore,the excess green index and the color index were both calculated for image segmentation.The genetic algorithm based on the maximum entropy was used to estimate the grassland vegetation cover from the segmented UAV imagery.Finally,the field survey data was used to assess the accuracy of estimated vegetation coverage based on the t-test.The results showed that the maximum entropy-genetic model using excess green index from UAV imagery achieved high accuracy for grassland coverage estimation with P-value=0.853 6,RMSE=6.635 6.The biomass model based on variables of both height and vegetation cover performed better with accuracy of 83.41%,R2 of 0.853 6 and RMSE of 2.332 0 g·m?2 hat based on a single variable of either height or vegetation.The research reveals that UAV imagery with maximum entropy-genetic model can not only provide quick,accurate,efficient ground data for vegetation monitoring,but also ground verification for vegetation estimation from remote sensing data.
作者
刘艳慧
蔡宗磊
包妮沙
刘善军
LIU Yanhui;CAI Zonglei;BAO Nisha;LIU Shanjun(Institute for Geo-informatics&Digital Mine Research,Northeastern University,Shenyang 110819,China;Survey and Design Institute of Water Conservancy and Hydropower,Jilin province,Changchun 130012,China)
出处
《生态环境学报》
CSCD
北大核心
2018年第11期2023-2032,共10页
Ecology and Environmental Sciences
基金
国家自然科学基金青年基金项目(41401233)
中央高校基本科研专项资金项目(N160102001)
作者简介
刘艳慧(1994年生),女,硕士研究生,主要从事植被遥感方面的研究。E-mail:LYH2513178156@126.com;通信作者:包妮沙(1985年生),女,副教授,博士,主要从事环境遥感、矿山土地复垦(矿区环境遥感监测)方面的研究。E-mail:baonisha@126.com。
