Plant Leaf Identification Using Color and Multi-scale Fractal Dimension

Author(s)

Backes, Andre R. ; Bruno, Odemir M. ; Elmoataz, A ; Lezoray, O ; Nouboud, F ; Mammass, D ; Meunier, J

• Document type: Journal article
• Source: Lecture Notes in Computer Science; v. 6134, p. 3-pg., 2010-01-01.

Abstract

The most extracted measures from plants are traditionally performed manually. This and the great biodiversity of species makes the taxonomical classification of plants a very complex and time-consuming task. In order to contribute to the plant species characterization, a novel approach based on color texture is proposed. Each color channel in a leaf texture is modeled as a surface, so the complexity in each channel, as well as in the interaction among these three color channels, can be measured using the Multi-scale fractal dimension. Results show the potential of this approach, which overcomes traditional color texture analysis methods, such as Chromaticity Moments and Gabor EEE descriptors. (AU)