As a TU Munich spin-off, OCELL uses the most modern computer vision algorithms in order to analyze tree stands automatically from aerial imagery and 3D point clouds. This saves time and money in your stock takin.
Additional to high res aerial imagery of your tree stands, we provide a complete GIS integration. With us, you have a partner for the complete process.
While stock taking is often only being done for tax reasons, with OCELL you receive regular updates on highly relevant information in order to improve your business processes.
With our airplane-based image capture we can offer ground sampling distances of 5-8 cm even for large areas like 10,000 hectars. Using modern lossless compression there's no need to be afraid of large amounts of data or slow download speeds. Additonally we offer a cloud-hosted access. Additional to RGB and NIR, we can also offer further spectral channels on request such as certain bands for vitality research.
Conventional computer vision algorithms all failed in delivering reliable classification of different tree species. Together with the TU Munich Chair of Data Processing, OCELL developped special AI-driven computer vision algorithms, which manage to discriminate between five different tree species reliably and provably. This way not only pure pine or spruce stands can be analyzed but also mixed coniferous stands.
(Move slider to compare)
A problem of every forest manager can now be solved quickly and cost-efficient with OCELL analyses. Understories and thinly timbered stands are being discriminated through their height information and analyzed separately. This way the stock of the top-stand can be calculated precisely, without having the understory distort the results. At the same time the height of the understory can be measured.
Through the use of localized harvester data, indivudual sub-areas can be associated with stock changes. OCELL developed a model from the wood volume measurements of the harvesters, which allows an extrapolation to all tress of the stand. With this method we can make more precise statements about stock and growth than with yield tables.
You send us the border files of your forest areas, the results of previous forest stock-takings and harvester data, if available.
Depending on demand we capture up-to-date high-res imagery of your areas up to 6 times a year as a foundation for all following analytics.
After the making of the initial stock taking, additional data from harvesters, or terrestrial measurements can continuously be integrated, in order to keep track of stocks and stock changes. The appearance and growth of calamities can also be monitored this way. All data can be displayed in QGIS or ArcGIS in a georeferenced way and will be updated with each OCELL upload.