Leaf Recognition and Classification Using GLCM and Hierarchical Centroid Based Technique

Variety of plants exists in earth's ecology where each plant species has its own unique features. Due to their immense benefits to mankind, many plant species are used in day to day life. Therefore, accurate plant leaf recognition through computer vision methods has paved its way to several fields like ayurvedic and diagnosis of health issues. Technology has always played a vital role in all aspects of human development. Achieving accurate recognition and categorization of plant leaf is always a challenge to researchers. This paper, therefore aims to put forth various techniques that are adopted for pre-processing, feature extraction and classification of leaf, based on shape and texture features of leaf sample. The paper further presents experimental results carried out on flavia dataset in order to recognize and classify leaves using gray level co-occurrence matrix and hierarchical centroid based technique. This work has considered 300 leaf samples with 30 different classes for the purpose of investigation. Experimental results indicate that from the aforementioned technique, it is possible to attain accuracy up to 96.66%.