The application of modern technologies makes it easy to collect, process, present, and apply data for logistics in hard to reach areas. Proper delivery of equipment, personnel, and materials directly affects the quality of work. The need for precise and real information about the condition of the terrain and the conditions of the environment has always existed since this knowledge enables proper planning, forecasting and task performing in the field. Improvement in the digital industry enables fast and easy transfer of unchanged digital data from the field to the information processing centers, which consequently improves decision making and planning processes. New workflows made proper logistics even more important because it increased the precision of field work and better anticipation of previously unforeseen circumstances. Work on hard to reach areas, with large slopes, non-existent and/or non-persistent infrastructure, and different degree of vegetation coverage requires precise planning and organization of works, in order to minimize the number of unforeseen situations and make the most expeditious workflows. This paper presents the practical application of small unmanned aerial systems for collecting a large amount of data in a short time, the processing of the data, and the production of relevant information for decision making. There are two most important aspects of this paper. First one is fast, easy, safe and precise collection of large amounts of data which is an alternative to the traditional methods. The second is computer data processing, which enables a fast and automatic transformation of raw data into relevant information in digital formats that are suitable for further processing and easily accessible to decision makers. This work shows that it is possible to record quickly and in detail a large area, and obtain real, current, accurate and high-fidelity information about each point of terrain, with high precision and reliability

In this paper, Soil Protection Coefficient (Xa) was quantified through the approach of high resolution multispectral orthomosaic segmentation and classification. The approach was presented in the example of ski lane in ski center Kopaonik. The data collection was performed through application of Unmanned Aerial System equipped with 5band multispectral sensor and RGB sensor. Data processing was performed with digital photogrammetric and Object Based Image Analyses software. The Soil Protection Coefficient represents the descriptive and very sensitive parameter of Erosion Potential Method. Application of 5 spectral bands, of which 2 bands are very sensitive to the type of land use /land cover allowed precise detection, delineation and classification of different land cover/use types. These types were directly tied to the values of Xa coefficient which were originally proposed by the author of the Erosion Potential Method, professor Slobodan Gavrilović. The final result was a georeferenced digital map classified with both land cover/use and Xa values classes. This approach created the potential to use such maps for further analyses, planning, and modeling of erosion protection measures.

In the area of Stara Planina mountain, a multispectral survey of forest vegetation was performed. Data acquisition was done with unmanned aerial system DJI Phantom 4 Pro, equipped with integrated RGB 20Mpix sensor, and MicaSense RedEdge M, 5-channel narrowband multispectral sensor. Data was collected in the form of images, and 4 composite orthomosaics were produced-broadband visible RGB, narrowband visible RGB, and with vegetation indices applied NDVI and NDRE. RGB orthomosaic showed no significant changes in canopies apart from the variability of levels of green. Orthomosaics with vegetation indices applied showed changes in the level of physiological activities of leaves. NDVI map showed the negative changes of the top of the canopies, while NDRE map showed more dramatic changes within the canopy as well. Besides the map, 5 polygons with different NDRE values were selected and their respective spectral signature graphs were produced. The areas with the lowest NDRE values had the highest reflectance values in all wavelengths, while the absorption of light is much higher in physiologically active vegetation. Values of NDRE lower than 0.479 were inspected from the ground. This kind of ground-truth provided evidence that the areas coded in red were with lower physiological activity due to the infestation by beech leaf-mining weevil Orchestes fagi L. Another interesting finding was that both NDVI and NDRE values were higher in the areas not directly exposed to the sunlight. The areas shaded by surrounding canopies received only diffuse light but it showed a more positive ratio between absorbed and reflected wavelength which could be characteristic of the Fagus Sylvatica species. The findings in this study showed a strong correlation between low values NDRE vegetation index and negative changes deep within the canopy of high vegetation, which can serve as an indicator of pest infestation in forestry.