Automatic leaf structure biometry: computer vision techniques and their applications in plant taxonomy

Abstract

This paper proposes a new methodology to extract biometric features of plant leaf structures. Combining computer vision techniques and plant taxonomy protocols, these methods are capable of identifying plant species. The biometric measurements are concentrated in leaf internal forms, specifically in the veination system. The methodology was validated with real cases of plant taxonomy, and eleven species of passion fruit of the genus Passiflora were used. The features extracted from the leaves were applied to the neural network system to perform the classification of species. The results showed to be very accurate in correctly differentiating among species with 97% of success. The computer vision methods developed can be used to assist taxonomists to perform biometric measurements in plant leaf structures.