| With the continuous development of unmanned aerial vehicle(UAV)technology,its application in forestry is becoming more and more widely.As an important way to obtain high score images,aerial photography of unmanned aerial vehicle has gradually become a new way of forest resource investigation and monitoring,with the characteristics of low cost,high efficiency and strong timeliness.In this study,the structure parameters of single tree and biomass inversion of urban forest demonstration base of Northeast Forestry University were studied by using small UAV with consumption-grade digital camera and terrain data of total station.The accuracy of the results was validated by the measured forest data of tree height measuring instrument,etc.The main results of this study include:(1)Tree Height Extraction Based on UAV and Total StationAiming at the problem that a single UAV image can not effectively extract the tree height for stands with high canopy closure,a method of obtaining the tree height based on the UAV image data and the terrain data from total station measurement was proposed.Fifty-five individual tree heights were extracted automatically in the study sample plot,and the extraction rate of tree number reached 91.7%.The automatic acquisition of tree height was realized.Compared with the actual tree height,the extraction accuracy of individual tree height reached 89.9%.(2)Extraction of canopy area of middle and high canopy density standAiming at the problem that the canopy boundary and area can not be accurately extracted due to the interlocking of tree crowns in the UAVs images of high canopy closure,a method of extracting canopy boundary and area is proposed by using the improved Hough transform algorithm while treating canopy boundary as a circle.47 and 26 crowns in coniferous forest and broad-leaved forest was automatically extracted while the extraction rate of the trees was 91.2%.Compared with the actual canopy area,the accuracy of canopy area extraction in coniferous forest and broad-leaved forest was 79%and 83%.(3)Forest Biomass Inversion and MappingAccurate determination of forest biomass is important for global climate change and carbon cycle research.Four sample plots of 20×20m were selected in the study area.The tree height and DBH data of each sample plot were measured by wood gauge.The H-DBH(tree height-DBH)estimation model was established.The R2of the four sample plots was greater than 0.7.The forest biomass of the forest farm was measured.The experimental results show that the average aboveground biomass was 6915.85 kg,and the average extraction accuracy was 88.6%.The above-ground biomass distribution map for the entire study area can be quickly obtained by Arcgis software combined with the research method proposed in this study.In summary,small UAV with digital camera platform can extract forest structural parameters of middle and high canopy density in a small area,and its accuracy meets the accuracy requirements of forest survey.The results of the study lay a foundation for acquiring basic information of tree parameters quickly,accurately,efficiently and detecting the dynamic changes of forest resources in time.Thus,the degree of digitalization and automation of forestry survey will be increased,and the measuring period will be reduced.The theory and application of digital forestry in China will be raised to a new level. |
PDF Full Text Request