February 2008


Computer Automated Detection and Measurement of Single Trees and Forest Inventory Parameters using Satellite and Airborne Colour Infrared Photos (CIR) and Laser Scanning Systems (LiDAR).

An illustrated digital version in print quality can be downloaded from http://landConsult.de/segmentation/download


At landConsult.de we specialize in processing digital aerial and satellite imagery of forest stands. These days the classical image processing methodology has been amended by new ways to analyse Airborne Laser Scanning data and High Resolution Imagery.
We developed an algorithm to identify single trees from colour infrared and high resolution airborne (e.g. Vexel UltraCam) or satellite (Quickbird, Ikonos) imagery. The program generates a shape file (GIS) representing the area and centroid of the detected tree crowns, including the tree height if a surface model is available. The tree height is processed from Laser Scanning (LIDAR) data or photogrammetrically from aerial or satellite (e.g. CartoSat) stereo models.

In addition landConsult and its partners, the University of Freiburg and Cracow, are permanently improving existing methodologies to filter and process huge laser point clouds to automatically measure all laser-represented objects, such as the forest road network, forest stands or single trees. By means of scientifically proven correlations between the number of stems per ha, the crown area and the tree height, the standing stock volume can be assessed. The discrimination of different tree species is being done by a combination of image and LiDAR processing.



Our Service:

Geographical recording (x/y-coordinates) of all dominant and visible sub-dominant trees from aerial imagery and LiDAR or from space (Quickbird).

Calculation of tree height (z-coordinate) of all detected trees.

Visualisation and classification (species, height, density) of all detected trees in a GIS (point shape file).

Classification of forest stands in terms of common features such as vertical and horizontal stand structure, stand density, crown closure, mean stand height, highest trees in a stand, and distribution of tree species.

Assessment of standing stock volume per hectare and per tree species, calculation of single tree volume.

Generation of a high resolution terrain model representing slope, aspect and hidden geo-morphological features covered by vegetation (e.g. ditches, rocks, death wood, etc.).

Automatic detection and classification of roads, logging- and temporary tending lines.

Design of a reliable and cost effective Forest Monitoring System based on Satellite data.

Presentation of results in an Orthophoto Map and GIS, if required with authorised access on a web based system (GIS MapServer or GoogleEarth).