Multi-phase ALS Point Clouds Registration And Growth Monitoring Of Plantation

Accurate monitoring of plantation resources is the premise for improving the quality of plantation cultivation,sustainable management and accurate estimation of plantation carbon storage.Using airborne lidar can obtain high-precision forest canopy structure information,and obtaining multi-phase lidar point clouds in the same forest area is the basis for effective dynamic monitoring of plantation resources.However,because lidar equipment is a precision instrument,it is affected by systematic errors,weather,etc.during use,which will inevitably lead to deviations in the coordinates of point clouds obtained in the same forest area in different periods.After the point clouds is accurately registered,the dynamic monitoring of plantation resources can be accurately realized.At the same time,with the rapid development of UAV equipment,UAV-borne Li DAR technology is becoming more and more mature.The use of UAV-borne lidar to collect forest data is inexpensive and easy to operate,and the obtained point cloud density is high,and higher-precision forest structure information can be obtained,which greatly reduces the difficulty and cost of obtaining plantation resource data.However,after the data is obtained,the dynamic monitoring of plantation resources is still mainly carried out by manpower,which affects the overall process of plantation resource monitoring.Therefore,the development of a resource monitoring system suitable for plantation can effectively improve the efficiency of plantation data processing,speed up the process of dynamic monitoring of plantation resources,and is of great significance for improving the quality of plantation cultivation and sustainable management.In view of this,the main innovations and contributions of this thesis are as follows:(1)Aiming at the characteristics of single species,regular arrangement,and lack of texture features in the plantation,based on the relative spatial relationship of the trees,a highly robust multi-phase airborne lidar plantation point clouds matching algorithm was created: first,using The ground points are registered on the z-axis,and the point clouds of the two periods are segmented to obtain the tree position and height information,and the individual tree matching features are extracted according to the relative relationship between the horizontal and vertical directions of the trees;secondly,an appropriate similarity function is established,combined with the individual tree matching feature to construct a weighted bipartite graph model,and use the maximum weight matching algorithm to obtain the corresponding relationship between the two periods plantation;finally,use the singular value decomposition to solve the optimal transformation matrix to complete the registration.After the registration is completed,the two periods plantation are repaired according to the set distance threshold,and the tree height change is calculated according to the individual tree height information.Through the experimental verification in the typical coastal artificial forest research area of Jiangsu Province(the main tree species are poplar and Metasequoia),the results show that the created matching algorithm has a good registration effect in the typical sample plots of Metasequoia and Poplar,and the registration accuracy meets the requirements of the forestry,the registration result of the Metasequoia plot(RMSE=42.5cm after registration)is better than that of the poplar plot(RMSE=58.8cm after registration).The algorithm can effectively improve the matching accuracy of multi-phase airborne lidar plantation point clouds.(2)Using Pycharm as the development platform,using Python language,Py Qt5 toolkit and Open3 d toolkit to develop a point clouds matching and change monitoring system for plantation.After feasibility analysis,requirement analysis and overall design of the system,First,a data processing module is designed,including point cloud denoising and coordinate transformation functions,which can perform basic data preprocessing on the acquired plantation point cloud data;secondly,a point cloud visualization function is designed,which can analyze the data processing results and registration results;then a point cloud registration function is designed,and this function is used to complete the accurate registration of the two-phase plantation point clouds.The monitoring function analyzes the tree height change of two plantations.After testing,the system functions well and meets the needs of plantation forest resource monitoring.The system can effectively improve the monitoring efficiency of plantation resources,save manpower and material resources,and is of great significance for scientific management and cultivation of plantations.