Signature of plant leaf [electronic resource].

Includes reference.

This study aimed at overcoming some of the leaf shape analysis limitations by developing a robust method for plant identification depending on the morphology of leaf venation web. The specific objectives were to: 1. Develop preprocessing procedures to get color leaf images ready for extracting accumte morphological features. 2. Get leaf signature by calculating number of matches of a set of twenty selected morphological features. 3. Develop leaf signature prototypes and decision functions to build, train, and test a minimum distance (MD) classifier in identifying plants by classifying leaves: from three different plants (Alfalfa, Mint, and Pomegmnate), with two occlusion mtios (0%, 50%), at three axle directions (one horizontal and two diagonals), in other words, at three different directions of leaf orientation angles (0°, 45°, 135°), and of three different sizes (large, medium and small).

Summary in Arabic.

1