Representation and classification of medicinal plants: A symbolic approach based on fuzzy inference technique

Abstract

In this paper, a method of representing shape of medicinal plant leaves in terms of interval-valued type symbolic features is proposed. Axis of least inertia of a shape and the fuzzy equilateral triangle membership function is exploited to extract features for shape representation. Multiple class representatives are used to handle intra class variations in each species and the concept of clustering is used to choose multiple class representatives. A simple nearest neighbor classifier is used to perform the task of classification. Experiments are conducted on the standard flavia leaf dataset to demonstrate the efficacy of the proposed representation scheme in classifying medicinal plant leaves. Results of the experiments have shown that the method is effective and has achieved significant improvement in classification accuracy when compared to the contemporary work related to leaf classification.