Information Acquisition of Forest Resources Using Photographing from UAV: Case Study in the Mie University Forest, Hirakura

In recent years, unmanned aerial vehicles (UAVs), which have been internationally developed and utilized in many fields, has also attracted people’s attention in the forestry service. Using UAV systems as a data acquisition instrument and a monitoring tool, it can be expected to solve some issues in conventional forest resource inventories and enabling more efficient research. This study was conducted using structure from motion (SfM), which is a photogrammetric range imaging technique. This is accomplished using aerial photography with a UAV in the University forest and consists of: 1) generation of a 3D model of photographed objects, 2) estimation of tree height and DBH, 3) evaluation of the estimation accuracy, and 4) comparisons to measurement errors between two aerial photography sets. The result showed that root mean squared error (RMSE) in the estimated tree height was 1.58 m and RMSE in the estimated DBH from a relational expression between tree height and DBH was 3.88 cm, and RMSE in the estimated tree height between the first and second photography was 0.21 m. Keyword: forest monitoring, GIS, remote sensing, SfM, UAV