A Hierarchical Approach to Fine-Grained Visual Classification for Plant Image Analysis

Ram Jitendrabhai Zaveri, Matthew Keaton, Cole Henderson, Meghana Kovur, and Dr. Gianfranco Doretto, Benjamin M. Statler College of Engineering and Mineral Resources, West Virginia University, 1306 Evansdale Dr, Morgantown, WV 26506

Object classification from images is a standard problem in computer vision. State-of-the-art techniques are based on deep learning, a subfield of machine learning. While they perform well for generic object classification such as determining whether there is a cat or a dog in an image, there are application domains where they still fall short. One of them is the automated classification of plant species. This is a case where different plants might sometimes look very similar in pictures, while the same plant can appear very differently, because of the concurrent effects involved in the image formation process, which involve the shape of the scene, its material properties, the illumination conditions, and the viewpoint. We regard this scenario as a fine-grained visual classification problem “in-the-wild,” since no restrictions on the imaging conditions are imposed, which makes the task at hand especially challenging. To address it, we have developed an approach based on state-of-the-art deep learning techniques to identify the hierarchy of plant organs appearing in an image. In order to test the approach, we have curated a new large-scale plant image dataset. We developed a tool for annotating plant images, which we used for labeling the plant organs. Our approach and the dataset will serve as a benchmark for future developments for image species classification in-the-wild.

Additional Abstract Information

Presenter: Ram Zaveri

Institution: West Virginia University

Type: Oral

Subject: Computer Science

Status: Approved

Time and Location

Session: Oral 9
Date/Time: Wed 12:00pm-1:00pm
Session Number: 914
List other presenters in this same room and session