Mangroves are important marine ecosystems, and their canopy structure parameters are important bases for evaluating the growth status and carbon sequestration capacity of mangroves. Currently, mangrove canopy parameters are mainly acquired using airborne LiDAR. The main drawbacks of airborne LiDAR are its high cost and limited accessibility. Meanwhile, the application of UAV oblique photogrammetry in this field is relatively limited. Therefore, the mangrove restoration area in Hainan was taken as the research object. A mangrove instance segmentation dataset was independently constructed. The YOLACT instance segmentation model was innovatively applied to the estimation of mangrove canopy structure. An efficient monitoring scheme for mangrove canopy structure was established. Comparative experiments were carried out with traditional methods including watershed segmentation, point cloud distance discrimination clustering, and layer stacking. The results show that UAV oblique images can realize fine identification of individual trees in the mangrove restoration area. The overall accuracy of the four segmentation algorithms is all greater than 0.80, with F-scores around 0.90. Among them, the YOLACT model performs the best (OA=0.93, F=0.96). In the experiments of estimating canopy structure parameters, the YOLACT model exhibits the optimal fitting effect and the smallest error, and the mean absolute error percentage was reduced by approximately 2%~12% compared with other algorithms. Overall, the YOLACT model adopted in this study shows higher reliability and application potential in individual mangrove tree identification and canopy structure estimation. This provides a technical reference for mangrove monitoring and management.
无人机技术的快速发展,为红树林林分层面的物种分类、精细结构解析与单木尺度监测提供了有力支撑[11]。当前,基于无人机摄影测量技术和激光雷达(light detection and ranging, LiDAR)技术,已成为获取红树林垂直结构特征的两种主要技术手段[12]。在相关研究中,Yin等[13]将激光雷达数据获取的红树林树高与实地测量结果进行了比较,发现偏差为-9~11 cm;王德智[14]通过无人机搭载LiDAR获取了红树林冠层高度模型(canopy height model, CHM),为后续联合卫星遥感影像实现大面积红树林地上生物量反演奠定了基础。雷鹏[15]以无人机激光雷达数据为基础,采用分水岭分割算法、区域增长算法、最近邻算法、区域卷积神经网络算法进行单木分割,探究不同算法下红树林识别与分割效果。虽然机载LiDAR在红树林垂直信息识别上具有很大潜力,但是高额的成本与庞大的数据量阻碍了其在红树林调查中的进一步应用。相比之下,无人机摄影测量凭借成本较低、数据处理相对简单等优势,成为机载LiDAR技术的有效补充,受到了研究人员的重点关注[16]。Navarro等[17]使用基于局部最大值的可变窗口滤波来检测树顶,采用标记控制分水岭分割法估测了天然红树林和人工红树林的高度、树冠直径和地上生物量。余楚滢等[18]以无人机倾斜摄影生成的冠层高度模型为基础,采用种子区域生长算法进行无瓣海桑单木树冠提取。
综上,相较于机载LiDAR技术的相关研究,利用无人机倾斜摄影测量开展红树林冠层结构参数提取的研究仍存在诸多空白,不仅相关研究内容有限,针对不同算法在该场景下的识别效果、分割精度也缺乏系统性的对比分析与科学验证。针对上述研究短板,现以海南东寨港红树林修复区为研究区域,基于无人机倾斜摄影数据自主构建针对红树林的实例分割数据集,并将YOLACT(you only look at coefficients)实例分割模型创新应用于红树林的单木识别,与多种传统算法开展对比实验,探究不同算法的适用性与精度差异,为研究对象的监测与管理工作提供技术参考。
LymburnerL, BuntingP, LucasR, et al. Mapping the multi-decadal mangrove dynamics of the Australian coastline[J]. Remote Sensing of Environment, 2020, 238: 111185.
[2]
ChenS, ChenB, ChenG, et al. Higher soil organic carbon sequestration potential at a rehabilitated mangrove comprised of Aegiceras corniculatum compared to Kandelia obovata[J]. Science of the Total Environment, 2021, 752: 142279.
JiaMingming. Remote sensing analysis of mangrove dynamic changes in China from 1973 to 2013[D]. Beijing: Graduate University of Chinese Academy of Sciences (Northeast Institute of Geography and Agroecology), 2014.
ChenYanli, SunMing, ChenCheng, et al. Remote sensing identification of mangrove diseases and insect pests based on GF-1 satellite[J]. Science Technology and Engineering, 2023, 23(34): 14682-14690.
LeiWenzheng, LuoTianyu, GuoXi, et al. Classification of mangrove tidal flats in the intertidal zone of Guangxi coastal zone based on GEE[J]. Science Technology and Engineering, 2025, 25(18): 7493-7501.
WangWenqing, ZhangLin, ZhangYamian, et al. Research progress on mangrove restoration through pond reversion[J]. Journal of Xiamen University (Natural Science Edition), 2021, 60(2): 348-354.
[11]
LucasR, van De KerchoveR, OteroV, et al. Structural characterisation of mangrove forests achieved through combining multiple sources of remote sensing data[J]. Remote Sensing of Environment, 2020, 237: 111543.
[12]
YangY, LiangX Q, WangB, et al. Biophysical parameters retrieval of mangrove ecosystem using 3D point cloud descriptions from UAV photographs[J]. Ecological Informatics, 2022, 72: 101845.
LiWenbo, LiJihong. Individual tree segmentation in high canopy density plantations using UAV high-density point clouds[J]. Journal of Northeast Forestry University, 2023, 51(9): 76-82.
[15]
XiaQ, QinC Z, LiH, et al. Evaluation of submerged mangrove recognition index using multi-tidal remote sensing data[J]. Ecological Indicators, 2020, 113: 106196.
[16]
ShendrykY, VilasM P. Correction: detecting mangrove seedlings from UAV imagery using deep learning for restoration monitoring[J]. Scientific Reports, 2026, 16: 4590.
[17]
OteroV, Van De KerchoveR, SatyanarayanaB, et al. Managing mangrove forest from the sky: forest inventory using field data and UAV imagery in the Matang mangrove forest reserve, peninsular Malaysia[J]. Forest Ecology and Management, 2018, 411: 35-45.
[18]
YinD M, WangL, LuY, et al. Mangrove tree height growth monitoring from multi-temporal UAV-LiDAR[J]. Remote Sensing of Environment, 2024, 303: 114002.
WangDezhi. Research on multi-scale observation methods of mangroves combining UAV-LiDAR and satellite remote sensing data[D]. Wuhan: China University of Geosciences, 2020.
[21]
雷鹏. 基于无人机激光雷达的红树林单木结构参数估测研究[D]. 桂林: 桂林理工大学, 2022.
[22]
LeiPeng. Estimation of individual mangrove structural parameters based on UAV LiDAR[D]. Guilin: Guilin University of Technology, 2022.
[23]
YeN, van LeeuwenL, NyktasP. Analyzing the potential of UAV point cloud as input in quantitative structure modelling for assessment of woody biomass of single trees[J]. International Journal of Applied Earth Observation and Geoinformation, 2019, 81: 47-57.
[24]
NavarroA, YoungM, AllanB, et al. The application of unmanned aerial vehicles (UAVs) to estimate above-ground biomass of mangrove ecosystems[J]. Remote Sensing of Environment, 2020, 242: 111747.
YuChuying, GongHui, CaoJingjing, et al. Individual tree extraction and above-ground biomass estimation of Sonneratia apetala based on UAV imagery[J]. Tropical Geography, 2023, 43(1): 12-22.
LiYuanling, ChenZongzhu, LeiJinrui, et al. Study on plant diversity of mangrove wetland in Dongzhaigang, Haikou[J]. Tropical Forestry, 2019, 47(2): 48-51.
[29]
邓湘娟. 东寨港红树林生态系统服务价值变化及提升方案[D]. 长沙: 湖南农业大学, 2018.
[30]
DengXiangjuan. Changes in ecosystem service value and improvement plans of mangroves in Dongzhaigang[D]. Changsha: Hunan Agricultural University, 2018.
TianMi, ZhongCairong, FangZanshan, et al. Evolution analysis of mangroves under different pond reversion modes in Dongzhaigang based on multi-source remote sensing[J]. Wetland Science & Management, 2023, 19(3): 4-8.
[33]
BolyaD, ZhouC, XiaoF Y, et al. YOLACT: real-time Instance Segmentation[C]// Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV). Seoul: IEEE, 2019: 9157-9166.
[34]
BiswasH, ZhangK, RossM S, et al. Delineation of tree patches in a mangrove-marsh transition zone by watershed segmentation of aerial photographs[J]. Remote Sensing, 2020, 12(13): 2086.
YueAnzhi, YangJianyu, ZhangChao, et al. Segmentation of high-resolution remote sensing images based on multi-scale morphological filtering[J]. Transactions of the Chinese Society of Agricultural Engineering, 2013, 29(S1): 89-95.
[37]
LiW K, GuoQ H, JakubowskiM K, et al. A new method for segmenting individual trees from the lidar point cloud[J]. Photogrammetric Engineering and Remote Sensing, 2012, 78(1): 75-84.
[38]
AyreyE, FraverS, KershawJ A J, et al. Layer stacking: a novel algorithm for individual forest tree segmentation from LiDAR point clouds[J]. Canadian Journal of Remote Sensing, 2017, 43(1): 16-27.
ZhuBodong, LuoHongbin, JinJing, et al. Optimization of individual tree segmentation method for high canopydensity plantations using UAV LiDAR[J]. Scientia Silvae Sinicae, 2022, 58(9): 48-59.
ZhangJingrui, RenGuangbo, WuPeiqiang, et al. Extraction method of individual mangrove growth parameters by combining airborne and terrestrial LiDAR[J]. Journal of Remote Sensing, 2025, 29(3): 728-740.
HuangXinxi, XiaKai, FengHailin, et al. Detection and segmentation of individual tree crowns based on UAV imagery and Mask R-CNN[J]. Forestry Science and Technology Development, 2021, 6(2): 133-140.
[45]
GessicaA, AngelaM, VittoriaP, et al. Use of high-resolution multispectral UAVs to calculate projected ground area in Corylus avellana L. tree orchard[J]. Sensors, 2022, 22(19): 7103.
[46]
ZhouL, MengR, TanY, et al. Comparison of UAV-based LiDAR and digital aerial photogrammetry for measuring crown-level canopy height in the urban environment[J]. Urban Forestry & Urban Greening, 2022, 69: 127489.
WeiXiaobo, LiZhiqiang, ChenZhaoguang, et al. Individual tree segmentation of different mangrove species based on UAV lidar[J]. Science Technology and Engineering, 2024, 24(3): 963-969.