Research On The Method Of Obtaining Tree Measurment Factors By Terrestrial Close Range Photogrammetry

Forest resources' monitoring and investigation are the data support and important basis for forest resources managcinent and operation in China,and observation instruments and observation methods are the important factors that determine the efficiency and quality of forest resources survey.With the continuous maturity of close-range photogrammetry technology.it has changed the traditional means of forest survey,and promoted the digital and intelligent development process of forest survey.Based on the ground close-range photourammetry technoloogy,in order to improve the accuracy of tree-rneasurement factors and reduce the working strength of field operation,the following three observation schemes are studied and proposed to observe individual tree or the sample plots of standing forest and to extract relevant tree-measurement factors.(1)This research proposes a set of technology method of tree measurement using monolithic photography of smart phone,and develops tlie corresponding system.In the process of forest observation,it is only necessary to take a photograph to calculate the DBH.tree height,volume,average DBH of standing forest,density,and stock volume of individual tree.The experimental results show that the relative deviation ot DBH of individual tree measured by this system is 4.95%and the relative root-mean-square error is 6.09%;The relative deviation of tree height of individual tree measured by this system is 5.59%and the relative root-mean-square error is 6.43%;The relative deviation of volume of individual tree measured by this system is 5.86%,and the relative root-mean-square error is 7.46%:The relative deviation of average DBH of standing forest calculated by this system is 0.51%,and the relative root-mean-square error is 8.6%;The relative deviation of density of standing forest calculated by this system is 1.46%,and the relative root-mean-square error is 7.47%;The relative deviation of stock volule of standing forest calculated by this system is 1.35%,and the relative root-mean-square error is 7.74%?All the above mentioned data all meet the requirements of deviation accuracy of forest survey in China.so the system has good prospects for application.(2)This paper researches a method for extracting tree measurement factors from individual trees(isolated trees)based on the close-range photogrammetry method of monocular vision.By reconstructing the three-dimensional point cloud lodel and combining with the established algorithm.this method extracts tree height,diameter of tree trunk at any height,average crown width and other tree measurement factors.The results of precision analysis of experiment show that the relative deviation of the extraction of DBH is 2.23%,and the relative root-mean-square error is 6.99%;the relative deviation of the extraction of individual tree height is 3.29%,and the relative root-mean-square error is 9.77%'The rclative deviation of the extraction of average crown width was 8.11%,and the relative root-mean-square error was 9.57%.All above mentioned accuracy can meet the needs of forest survey in China.(3)This research proposes a method to extract the DBH of trees in sample plots of standing forest by making use of the camera to shoot video,observing the sample plots,and establishing three-dimensional point cloud.This method is not only simple and feasible,but also limited to very few environmental factors by attempting to apply video measurement technology to forestry survey.A new data processing method of point cloud is proposed according to the characteristics of point cloud model and the fact that there will be shelter in sample plot.It extracts the slices of point cloud at DBH,and extracts the DBH of trees by cylindrical fitting and circle fitting.Finally,the experimental verification results show that,the relative deviation of the precision evaluation parameters of the method of extracting DBH by more effective and relatively accurate circle fitting is-1.49%?5.53%,and the relative root-mean-square error is 3.14%?8.61%.They all meet the requirements of deviation accuracy of forest resource survey in China.Therefore,this method can be popularized in the process of forestry investigation in China.Each of these three schemes has its own advantages and their precisions of observation can meet the demand of forestry survey in China,which has relative high realistic meaning and practical application value.