Tree species classification
Identify tree species remotely for ecological evaluation and enhanced forest management.
This solution allows you to identify the tree species based on the analysis of imagery retrieved from the Sentinel-2 satellite.
The identification/classification process is carried out within a selected area (AOI) based on the data collected over a considerable period of time (preferably from several months to one year).*
Key features of the solution:
-Accuracy of up to 80% with ground data available.
-High level of detail in tree species classification.
Accuracy of the satellite analytics greatly depends on the accuracy of ground truth data. We will need information about at least 100 trees/plots for each tree species to achieve precision of up to 80%.
For countries with trees commonly growing in the temperate climate conditions (e.g. Ukraine), the classification process usually takes up to two months. More “exotic” tree species will necessitate additional time for training the neural networks (up to 4 months), with mountain regions being especially tricky.
Key benefits of the solution:
-Results delivered in a variety of convenient formats (API, file sharing, email, and more).
-High level of customization to give you full control over the outcome.
And more.
*To ensure the highest accuracy possible we need a sufficient amount of cloudless images.
Other Solutions
Keep vigil of your forest cover by monitoring changes and accurately estimating area at a glance.
Take a precision care of the environment by remotely monitoring deforestation in selected areas of interest
Remotely track the success of your reforestation efforts