基于定位观测站图像实现草原植被覆盖度自动化提取的方法比较

Comparison of automatic extraction methods of vegetation cover based on the grassland positioning observation station

植被覆盖度是草原生态监测和研究中极其重要的指标。本研究采用无人值守定位观测站获取样方图像数据集,比较选取多种提取算法,旨在解决定位连续观测中不同样方的大量图像数据集如何自动化提取植被覆盖度指标的难题。本文使用绿度指数法、随机森林(RF)、支持向量机(SVM)和反向传播(BP)神经网络不同图像分割方法提取不同草地类型的覆盖度,并讨论绿度指数之类的阈值分割方法与机器学习方法的优劣、3种机器学习算法产生不同分类效果的原因以及覆盖度数值产生误差的主要原因。结果表明,机器学习算法可灵活解决定位观测站样方图像集的快速自动化提取植被覆盖度问题。绿度指数阈值分割方法应用于植被覆盖度的分割效果较差;RF算法在高寒草原的分割精度较高,SVM在温性草原和温性荒漠草原中分割效果较高,BP神经网络在高寒草甸的覆盖度提取中更有优势。本研究可为新时代草原生态监测的信息化和智能化监测设备研发提供重要的参考依据。

Vegetation cover is an important index in grassland ecological monitoring.In this study,positioning observation stations were used to obtain the image of the sampling plot,and vegetation cover extraction algorithms suitable for different grassland positioning observation stations were selected and compared.This was conducted with the aim of solving the problem of how to automatically extract vegetation cover indices from images of different plots in continuous positioning observation.Different image segmentation methods such as the greenness index method,random forest(RF),support vector machine(SVM),and back propagation(BP)neural networks were used to obtain the cover extraction results for different grassland types.The pros and cons of threshold segmentation methods such as the greenness index and machine learning methods,the reasons for the different classification effects of three machine learning algorithms,and the main reasons for the error of coverage value were discussed in this paper.The results have shown that the machine learning algorithm could be flexibly applied to the rapid automatic extraction of vegetation cover in the quadrat image of the positioning observation station.The greenness index threshold segmentation method applied to the segmentation of vegetation cover was relatively poor.RF algorithm has a higher level of accuracy in the segmentation of alpine grassland.SVM has a higher level of accuracy in temperate grassland and temperate desert grassland.The BP neural network has more advantages in the cover extraction of alpine meadow.This study can provide an important reference for the development of information and intelligent monitoring equipment for grassland ecological monitoring in the new era.