Estimating basic forest stand parameters from UAS images

Owing to their numerous benefits to humanity and other species, forests are considered the lungs of our planet. Unplanned use of forests causes the humanity face with a wide variety of critical issues such as climate destabilisation, air pollution, temperature instability, landslides, etc. This, of course, attaches importance to the sustainability of forests. Managing forests in conjunction with technological developments play a significant role in the continuity of the service provided by forests. A high-quality forest management depends heavily on the proper specification of forest characteristics, i.e. stand parameters. These parameters are used to produce forest management plans. In Turkey, stand parameters are generally obtained with terrestrial measurements. However, field measurements are neither cost- nor time-efficient in most cases, especially in large-scaled forests. The aim of this thesis is to estimate critical stand parameters (location, number, height, crown closure and crown diameter of trees) with point clouds extracted from Unmanned Aerial Systems (UAS). A novel polynomial fitting-based (PFB) approach was also proposed to estimate crown diameters of trees. The results of the proposed methodology were compared to the reference data and to the results of widely used segmentation-based approaches. All stand parameters were estimated in MATLAB environment. The results showed that the developed MATLAB script managed to estimate the stand parameters with a high accuracy and that the PFB methodology outperformed the other segmentation-based approaches in crown diameter estimation. The main conclusion drawn from this thesis is that using UASs enables the accurate estimation of stand parameters.