| Forest resources are one of the most important resources for human beings.For a long time,ithas provided sufficient material basis for people’s survival and development.At the same time,it ensures the balance of the natural carbon cycle.The survival and development of human beings cannot be separated from forests.Human beings can better play the function of forest by obtaining forest ecological parameters to master the trend of carbon.Forest biomass has been widely used in the field of forest ecology research as an important parameter for carbon storage,carbon cycle and evaluation of forest ecological functions.The appearance of LiDAR remote sensing technology has made up some shortcomings of traditional forestry survey in dynamic monitoring and other aspects,it also provides various possibilities for accurate and efficient estimation of forest biomass and other forest parameters.With the gradual rise of the application of LiDAR in forestry.According to the different platforms used,LiDAR methods can be roughly divided into spaceborne LiDAR,airborne LiDAR and ground-based LiDAR.Airborne LiDAR is carried by plane or UAV as a platform.Domestic and foreign scholars have made a lot of research achievements in the retrieval of airborne LiDAR forest parameters.Due to technical reasons,the traditional UAV-LiDAR has short endurance time and general radar performance.It is difficult to obtain large-scale and high-density point cloud data.The ability of laser to penetrate the canopy is weak,and it is difficult to obtain relatively complete canopy information and ground information.It is difficult to extract more accurate three-dimensional parameters(tree height,crown width,crown projection area and volume)of individual tree.With the development of the times and the advancement of technology,people continue to upgrade and improve UAV and radar equipment,providing technical support for obtaining high-density UAV-LiDAR point cloud data.In order to explore the application potential of UAV-LiDAR for individual tree segmentationand biomass estimation of urban artificial coniferous forest,the Pinus koraiensis,Larix gmelinii and Pinus sylvestris in six artificial coniferous forest plots of Forest Botanical Garden of Heilongjiang Province were taken as the research objects.The plot area was 0.06 hectares.The high-density point cloud data(2138pts/m~2)of the plot was obtained by UAV,and the UAV-LiDAR point cloud was denoised and filtered.Based on CHM raster data and UAV-LiDAR point cloud data,the individual tree segmentation is carried out respectively.Both of them have achieved good segmentation results.The overall segmentation accuracy is 0.85 and 0.86respectively,and the segmentation method based on UAV-LiDAR point cloud is slightly better.Using UAV-LiDAR point cloud data to extract tree height and crown width parameters and perform regression analysis.The results show that the tree height R~2 is about 0.88 and the crown width R~2 is about 0.72,indicating that the extracted parameters are relatively reliable.On this basis,relatively accurate crown projection area and volume parameters can be obtained.According to the allometric growth equation of the Pinus koraiensis,Larix gmelinii and Pinus sylvestris published by the State Forestry Administration,the individual tree biomass of the sample plot was calculated as the true value of individual tree biomass.The extracted three-dimensional structural parameters(tree height,crown width,crown projection area and volume)were used to fit the individual tree biomass model of the Pinus koraiensis,Larix gmelinii and Pinus sylvestris in this area.Through correlation analysis,the R~2 of the three tree species was about 0.8,indicating a high correlation.When constructing the biomass model,the DBH data of individual trees were not used as the relevant parameters,and the results could still get a good fitting degree.So,this study provides a new way for the application of UAV-LiDAR in the biomass inversion of large-scale artificial coniferous forests. |
PDF Full Text Request