摘要
针对单木识别研究中"局部最大值"算法因窗口大小设置不同而产生的单木漏识别与错识别问题,提出了联合"局部最大值"与"单木树冠结构分析"的单木识别算法。算法首先利用"局部最大值"获得候选单木;然后对候选单木树冠结构进行分析,提取树冠结构曲线;最后通过对树冠结构曲线判别,剔除、合并错识别与过识别单木,从而提高单木识别的精度。选取大兴安岭林区8个典型样地进行实验,以实测可见单木为参考,与窗口大小为1.0m、2.0m的两种"局部最大值"算法进行比较。结果表明,该算法8个样地整体F测度为90.45%,相比1.0m、2.0m窗口的"局部最大值"法F测度74.82%、77.35%,分别提高了15.63%、13.10%。
The traditional"local maximum"algorithm with different settings of window size could lead to the error and missing of individual tree recognition.This paper presents an algorithm combination of"local maximum"and "tree crown structure analysis"based on image matching point cloud.The algorithm obtained the candidate trees by"local maximum"and then got the crown structure curve by the analysis of the candidate tree.The false and over recognition trees were eliminated and merged by the crown structure curve.The precision of the algorithm was verified by 8 plots in Daxing'an mountains forest region.The visible trees were used as the reference.The results were compared with the results of the traditional local maxima algorithm with the window size by 1.0 mand 2.0 m.Results showed that F-Score of the proposed algorithm was 90.45%,which increased by 15.63%and13.10%from 74.82%and 77.35% when compared with the traditional local maxima algorithm with the window size by 1 mand 2 m,respectively.
引文
[1]OLSON J S,WATTS J A,ALLISON L J.Carbon in live vegetation of major world ecosystems[M].USA:Rep.ORNL,1983.
[2]庞勇,李增元.激光雷达技术及其在林业上的应用[J].林业科学,2005,41(3):129-136.
[3]李增元,刘清旺,庞勇.激光雷达森林参数反演研究进展[J].遥感学报,2016,20(5):1138-1150.
[4]HAO Z Q.Vertical structure and spatial associations of dominant tree species in an old-growth temperate forest[J].Forest Ecology&Management,2007,252(1/3):1-11.
[5]WANG C.Biomass allometric equations for 10co-occurring tree species in Chinese temperate forests[J].Forest Ecology&Management,2006,222(1):9-16.
[6]刘晓双,黄建文,鞠洪波.高空间分辨率遥感的单木树冠自动提取方法与应用.浙江农林大学学报,2010,27(1):126-133.
[7]WANG L,GONG P,BIGING G S.Individual tree-crown delineation and treetop detection in high-spatial-resolution aerial imagery[J].Photogrammetric Engineering&Remote Sensing,2004,70(3):351-357.
[8]NEVALAINEN O,HONKAVAARA E,TUOMINEN S,et al.Individual tree detection and classification with uav-based photogrammetric point clouds and hyperspectral imaging[J].Remote Sensing,2017,9(3):185.
[9]WANG Y S,HYYPPJ.International benchmarking of the individual tree detection methods for modeling 3-d canopy structure for silviculture and forest ecology using airborne laser scanning[J].IEEE Transactions on Geoscience and Remote Sensing,2016,54(9):1-17.
[10]KAARTINEN H,HYYPPJ,YU X,et al.An international comparison of individual tree detection and extraction using airborne laser scanning[J].Remote Sensing,2012,4(4):950-974.
[11]KATHURIA A,TURNER R,STONE C,et al.Development of an automated individual tree detection model using point cloud LiDAR data for accurate tree counts in a Pinus radiata plantation[J].Australian Forestry,2016,79(2):126-136.
[12]POPESCU S C,WYNNE R H.Estimating plot-level tree heights with lidar:local filtering with a canopy-height based variable window size[J].Computers and Electronics in Agriculture,2002,37(1/3):71-95.
[13]SWETNAM T L,FALK D A.Application of metabolic scaling theory to reduce error in local maxima tree segmentation from aerial LiDAR[J].Forest Ecology and Management,2014,323:158-167.
[14]DANDOIS J P,ELLIS E C.High spatial resolution three-dimensional mapping of vegetation spectral dynamics using computer vision.Remote Sensing of Environment,2013,136:259-276.
[15]刘清旺,李增元,陈尔学.利用机载激光雷达数据提取单株木树高和树冠[J].北京林业大学学报,2008,(6):83-89.
[16]FALKOWSKI M J,SMITH A M S,GESSLER P E,et al.The influence of conifer forest canopy cover on the accuracy of two individual tree measurement algorithms using lidar data[J].Canadian Journal of Remote Sensing,2008,34(S2):S338-S350.