Wood Species Recognition System

From an aerial view, trees may look simply ordinary, with little to distinguish individual species beyond a guess of their native origin. In 2016, Malaysian-based researchers developed a neural network capable of identifying tree species by classifying pixel data based on colour and texture with an accuracy rate of 97%. The current neural network depends on what is known as the Improved-Basic Gray Level Aura Matrix (I-BGLAM) technique, teaching the algorithm to identity minute differentials in greyscale between pixels and matching the colour concentration and overall texture to its training database. This database contains information on the aspects of 52 tree species based on 100 images collected not by satellite, but through aerial photography. The number of species used to build the neural network is small (considering there are thousands of tree species), but if tree species information and imagery were available for areas globally it could be applied in more forests outside of the those in Malaysia which acted as test samples. The application of this technology is meant to ensure that land parcels where logging is allowed do actually contain the tree species that a logging company is authorized to cut. Conversely, if illegal logging takes place, regulators could hypothetically use this information to check what trees existed in the area that was logged, and where it took place. The application now is limited to the test site, but in the future with more widespread usage, this technology could be a boon for preventing illegal resource extraction and to ensure more accurate logging.

Zamri, Mohd Iz’aan Paiz, Anis Salwa Mohd Khairuddin, Norrima Mokhtar, and Rubiyah Yusof. “Wood species recognition system based on improved basic grey level aura matrix as feature extractor.” Journal of Robotics, Networking and Artificial Life 3, no. 3 (2016): 140-143.


Artificial Intelligence, Illegal Resource Extraction, Industry/Natural Commodities, Visual Technologies