Unmanned Aerial Vehicle (UAV) images can be considered an alternative to traditional or more costly methods. In the field of forestry inventory, UAV can be used to measure different properties of forest such as Canopy Height Model and tree types. To estimate tree heights using UAV images, an accurate DSM and also DTM over the desired area are required. For this purpose, one of the most well-known computer vision-based methods is Structure from Motion (SfM). The algorithm uses Scale Invariant Feature Transform (SIFT) to extract and match point features. Then, these points are used as tie points to solve a bundle adjustment and generate a point cloud. The produced point cloud should be processed to generate a Digital Surface Model (DSM). By subtracting the DSM and a DTM, the resultant is an nDSM or tree heights. For this purpose, series of UAV images are acquired and the above mentioned procedure is done in detail and tested over the dataset.
