This database called LEAF was originally created for experiments with recognition of wood species based on a leaf shape (see Recognition of woods by shape of the leaf). It contains leaves of 90 wood species growing in the Czech Republic, both trees and bushes; native, invasive and imported (only those imported species which are common in parks are included). The number of samples (leaves) of one species varies from 2 to 25, their total number in the database is 795. The leaves were scanned with 300 dpi, thresholded (binarized) , preprocessed (denoising and cleaning) and saved in PNG format. Original color scans are not publically available. The name of each file includes the latin name of the species and the label of the sample.
The database can be used for any experiment with binary images. For non-profit and education purposes, it can be used free of charge. For the conditions of commercial usage, please contact Jan Flusser (firstname.lastname@example.org).
When using a database, cite it as
LEAF - Tree Leaf Database, Inst. of Information Theory and Automation ASCR, Prague, Czech Republic, http://zoi.utia.cas.cz/tree_leaves
An example - a leaf of "Acer pseudoplatanus" in Latin, "Sycamore maple" in English, "javor klen" in Czech, sample No. 7, 4x scaled down:
The file 'list of species.txt' contains the list of species. The number after the semicolon is the number of samples of the current species. The file 'Acces to image files.m' is MATLAB m-file with the code for the acces to the image files.
Contact: Tomas Suk (email@example.com)
|tree leaf database.zip||3.9 MB|
|list of species.txt||1.9 KB|
|Acces to image files.m (no cells)||1.17 KB|
|Acces_to_wood_files.m (cells)||1.04 KB|