REFERENCES

VARIOUS VEGETATION INDICES

Vegetation indices are multi-spectral indicators based on the spectral (light reflectance) properties of vegetation and combine information from several wavelength ranges. They make it possible to quantify the condition, density and vitality of vegetation in a way that cannot be achieved through direct field measurements alone. In forestry applications in particular, they enable the consistent, repeatable and time-comparable assessment of large areas, thereby supporting the early detection of damage and the continuous monitoring of stand condition. In addition to a conventional camera system operating in the visible (RGB) range, the drone technology used by DualWing Forestry is also capable of capturing imagery in several further bands, including red, red-edge and near-infrared wavelengths, from which various vegetation indices can be derived. Among these, NDVI is the most widely used, but other indices such as GNDVI, NDRE and LCI are also of considerable importance.

NDVI

What can it be used for?

Assessment of vegetation cover and general vitality
Delineation of damaged areas (drought, windthrow, clear-cutting)
Monitoring of temporal changes

Forestry advantages:

Fast, robust and easy to calculate
Ideal for broad overview analysis of large areas
Well suited as a baseline layer for further investigations

GNDVI

What can it be used for?

Klorofilltartalom becslése
Korai stressz (aszály, tápanyaghiány) kimutatása
Fiatal állományok vizsgálata

Forestry advantages:

More sensitive to physiological changes than NDVI
Indicates problems at an earlier stage
Particularly suitable for monitoring regeneration and young forest stands

NDRE

What can it be used for?

Analysis of dense, closed canopies
Detection of hidden stress
Assessment of chlorophyll and nitrogen status

Forestry advantages:

Does not saturate as quickly as NDVI
Sensitive to subtle canopy changes
Effective early indicator in cases of stress and damage

LCI

What can it be used for?

Detailed estimation of chlorophyll content
Detection of early stress and degradation
Identification of subtle differences in condition

Forestry advantages:

Highly sensitive to the physiological condition of the canopy
Enables more precise analysis even in closed stands
Well suited for detailed monitoring and research purposes

SURVEYS IN LEAF-ON CONDITIONS

Drone-based data acquisition carried out under leaf-on conditions enables the generation of a detailed three-dimensional point cloud and surface model from high-resolution aerial imagery, accurately representing the spatial structure of the forest canopy. This model allows the direct, spatially explicit assessment of structural characteristics of the forest stand.

During processing, the previously generated terrain model is subtracted from the surface model, making it possible to determine the actual height of the trees. Canopy height, calculated as the difference between the surface model and the terrain model, provides a reliable basis for analysing the vertical structure of the stand.

Based on the highly detailed model, crown surfaces associated with individual trees can also be delineated, enabling the examination of the spatial arrangement of the stand as well as the estimation of crown cover.

By linking remotely sensed data with ground-based reference measurements, indirect estimation of timber volume also becomes possible. Height and crown-size parameters can be statistically related to field sample plot data, including stem measurements, allowing reliable quantitative estimates to be produced at stand level.

This method can be applied with particularly high reliability in mature and overmature stands, where the canopy is already well developed and the structure of individual trees can be identified with confidence.

SURVEYS IN LEAF-OFF CONDITIONS

During the survey, drone-acquired aerial images with high overlap are processed photogrammetrically to generate a high-density point cloud, providing a detailed representation of both the forest stand and the spatial structure of the ground surface. The resulting point cloud contains information on the canopy, the stems and the terrain surface alike, thereby enabling complex forestry analyses.

As the next step in processing, a highly accurate terrain model is generated by separating vegetation from the ground surface. This is of key importance for forestry applications, since knowledge of the terrain surface makes it possible to determine the actual height of trees. The difference between the canopy model and the terrain model provides reliable tree-height data, which serve as a basis for volume, biomass and growth analyses.

The high-resolution point cloud is not only suitable for analysing height relationships; with appropriate methods, the position of individual trees can also be identified, making it possible to estimate stem number and to examine stand structure in detail. This is particularly useful in forest inventory, regeneration planning and long-term monitoring applications.

TIMBER YARD SURVEY

During the survey, drone-acquired aerial images with high overlap are processed to generate a high-density point cloud that provides a detailed description of the surveyed area and the spatial structure of the timber stacks. The resulting dataset makes it possible to characterise the geometry of the stored timber and to map the surface with high accuracy.

Based on the surface models generated during processing, the height relationships and spatial extent of the stacks can be determined. This is of key importance for the geometry-based quantitative assessment of timber stock, as the method relies on the spatial description of the stacks.

A detailed presentation of the methodology and its scientific foundation is provided in Fekete’s (2025) diploma thesis, which examines the forestry applications of modern drone-based surveying methods.