Wavelet and pyramid histogram features for image-based leaf detection

Abstract

Image recognition-based methods have been widely used in leaf detection. Leaf detection can play an important role in agriculture in detecting potential plant diseases, phenotyping and taxonomy. In this paper, we present an analysis on the wavelet and pyramid histogram-based features for leaf detection. We have used Haar wavelet transform and pyramid histograms from original image as feature extraction method. The experiments were done on a standard dataset of leaf images of 36 different species of plants. We have tested the effectiveness of different types of features over a large variety of supervised machine learning algorithms. We propose the use of Random Forest as the best performing classifier on this dataset using the selected features. Our proposed method achieved significantly better accuracy in comparison to the previous methods.