To achieve precise improvement of forest quality requires precise prediction and precise monitoring of forest growth and forest resources and good management of forests.
In recent years, UAV as a new type of technical equipment to obtain data sources, with its low cost, lightweight and flexible, convenient, safe, and fast access to high-resolution images to make up for the traditional satellite remote sensing image low resolution, easy to be affected by cloud cover data quality and re-entry cycle restrictions and other deficiencies. At the same time, UAV aerial photogrammetry technology can realize the data collection and acquisition of forest classes and small groups. The research object is no longer based on single trees and sample plots. Still, it can also take the whole forest as the research object, which makes the UAV technology play a great advantage in forest resource survey and dynamic monitoring.
UAV workflow for obtaining forest information
A UAV is an unmanned aerial device that forms an "unmanned aerial system" or "UAS" when combined with a ground control station and data link. The forest information is acquired based on the images taken by the UAS. Then the tree parameters are extracted after aberration correction, connection point extraction, leveling, air-triple encryption, automatic image matching, orthophoto correction, etc. "Unmanned Aerial System" has been proved to be an effective tool for forest resources survey, which can provide low-cost and high-precision remote sensing data for forestry. UAVs can obtain specific tree information and carry out targeted forest resource surveys and forest area planning.
How to obtain forest structure parameters
Forest structure parameters are important indicators reflecting forest quality and management status. In the past, the rough forest resources survey was often measured by visual inspection or simple instruments with poor accuracy, which could not meet the requirements of "precise forestry" and "digital forestry." With the decreasing forest resources and the continuous development of digital technology, the accurate and non-destructive measurement of single trees and forest stands has become a hot research topic in related fields. At present, the forest structure parameters extracted by UAV images mainly include single tree height, single tree crown width, plant density, depression, average stand height, biomass and storage volume, etc. In contrast, tree diameter at breast height is mainly inferred from tree height and crown width extracted by images. With the continuous development of UAV technology and image processing technology, the extraction of forest structure parameters from UAV images is developing in digitization, informatization, and automation, and the extraction results will have more advantages in terms of accuracy and speed.
Applications in forest area planning and spatial distribution of trees
The application of drones in forest resource survey and forest area planning is developing in synthesis and modernization. Researchers use aerial digital photogrammetry images for the automatic generation of forest area boundaries. The accuracy of the obtained meets the actual production needs of forestry, thus significantly reducing the workload of field surveys and improving work efficiency. Various thematic maps generated based on UAV images provide new ideas for forest management and management. Forest resources zoning analysis using software produced thematic maps of forest zoning based on UAV aerial images. The overall accuracy of field verification reached 88.7%, indicating that UAV remote sensing technology is feasible in forest area planning.
UAV images have a natural advantage in forest resource surveys and large-scale inventories. UAVs can obtain large-scale tree information at a low cost, and specific information such as tree species distribution can be obtained through post-processing of the images. Experiments have shown that spatial distribution surveys of major tree species using UAVs are similar to those of ground surveys. A researcher acquired 25.29 square kilometers of UAV images of the study area and used object-oriented classification methods to obtain the spatial distribution of mangrove trees in the study area. The validation showed that the accuracy of the information exceeded 90%. By carrying different types of sensors, it is possible to obtain spectral information in different wavelength bands, which greatly enhances UAVs' forest survey capability and efficiency. UAVs' continuous expansion plays an increasingly important role in urban planning, renovation, and greening with the continuous expansion of the urban scale. Some studies have shown that using UAV images to obtain the distribution of green land cover in Jinggangshan under the situation of poor timeliness of satellite images and cloud cover, and the obtained results meet the requirements of urban greening measurement work.
Complete 3D reconstruction of forest landscape
UAVs can perform 3D landscape reconstruction very well. For 3D models with low accuracy requirements, UAVs have the advantages of fast information acquisition, simple processing, and strong mobility. With the continuous development of information technology and computer technology, reconstructing models using remote sensing technology has gradually matured. The reconstruction of 3D models of tree canopies can be completed by acquiring the depth information of UAV images through 3D reconstruction technology. In contrast, active remote sensing techniques such as radar airborne platforms only provide partial canopy reconstruction, and ground-based LiDAR data have limitations and are costly. Researchers use unmanned aerial vehicles to fly slowly near the target standing tree according to a predetermined trajectory to obtain images and use computer vision methods to build three-dimensional models of single trees and forest stands. Then use the UAV image to extract the digital elevation model and superimpose the high-resolution UAV image, and then use the ground control points collected by the carrier phase difference technology to verify the accuracy of the extracted digital elevation model. In conclusion, the forest 3D landscape constructed by UAV better shows the forest value, which is beneficial to forest protection and management.
UAVs and related technologies have advanced the application of UAVs in forest information acquisition, and the needs of modern forestry investigation and planning have likewise increased the research efforts of UAVs in forestry. Whether as an image acquisition source or integrated LiDAR data, UAVs can be well applied to many aspects of forestry.
When using drones for surveys, different types of cameras and flying heights can be selected to obtain remote sensing images with different resolutions according to the different requirements for obtaining tree information. Studies have shown that images with different pixel resolutions significantly impact the accuracy of forest information acquisition.
Unmanned aerial vehicles are useful in forest resource surveys and forest landscape reconstruction, and forest area planning. In the future, strengthening the restoration of the three-dimensional landscape of important tourist areas by drones in forest eco-tourism areas can provide effective services for scenic spots and local tourism promotion.