Evaluation of the health state of forests and an effect of phosphite treatments with the use of photovoltaic SLE UAV
HESOFF project is being implemented in cooperation with the Institute of Aviation in Warsaw and the Forest Research Institute in Sękocin Stary and is co-financing from the European Commission and the National Fund for Environmental Protection and Water Management. The project is focused on the integration of innovative technologies with innovative methods of forest cultivation. Accordingly, we distinguish two basic objectives of the project:
- Evaluation of the influence of phosphites as elicitors of tree resistance to pathogens of the genus Phytophthora.
- Implementation of new methods of forest health state assessment and the effectiveness of cultivation through aerial imaging of the Unmanned Aerial Vehicle (UAV).
It is well known that remote sensing can provide data about the forest on a large and small scale, and even at a centimeter resolution about a single tree. However, efficient cooperation between the satellite service provider and a small research institution is quite difficult to implement. Similarly, we can say about local municipality receiving standard satellite images from the shelf. However, using Unmanned Aerial Vehicles (UAV), small research organizations can not only improve, but even harmonize and integrate their international activities in all areas of the European Union. Thanks to the autopilot function, the UAV has the ability to use the automatic take off and landing, and thanks to combinations: ground station – interface, human – device solutions to many challenges in the measurement and display of land and sea can be provided. Typically, a forest can be determined by its properties such as age distribution of trees, their species, density, substrate quality, etc. These characteristics can be obtained by visual interpretation of imaging acquired from UAV. After the completion of the HESOFF project and positive verification of its results, it would be advisable to develop software for automatic segmentation of the obtained forest data in order to minimize the manual intervention of the image analyst. The processing chain of several filtrations, segmentation and data connections could effectively and automatically produce homogeneous interpretation layers. The main sets of input data would be VHM and images from BSP. In addition, information on the existing infrastructure of forest roads and ducts, often analyzed as demarcation lines, could be included. Of course, not all forest properties can be obtained by visual remote sensing (eg soil characteristics). Nevertheless, such forest properties as spectral indexes of images that strongly correlate with tree species, tree height, tree structure and density are already possible to obtain this way.