Software that can identify a plant from an image of a leaf has been developed by researchers in Malaysia. The system uses statistical operators to reduce noise in images of different leaves to make the process more efficient.
Given the importance of plants for global ecology, it is important as humanity strives towards a sustainable approach to living that we have the right plants in place and that we recognize the rarities we best not lose. The team says their system could help in this regard. It extracts ten features from pre-processed leaf images and with one particular filtering approach, WFT (Wiener filtering technique) has an identification accuracy of 95.1%.
Aliyu, M.G., Kadir, M.F.A., Mamat, A.R. and Mohamad, M. (2018) ‘Noise removal using statistical operators for efficient leaf identification’, Int. J. Computer Aided Engineering and Technology, Vol. 10, No. 4, pp.364–377.