Project Team
Students
Alexander Petrov
Computer Science
Penn State Harrisburg
Meghaa Shanmugam
Biotechnology
Penn State Harrisburg
Faculty Mentors
Dr. Shobha Rudrabhatla
Penn State Harrisburg
Department of Biological Sciences
Dr. Sairam Rudrabhatla
Penn State Harrisburg
Department of Biological Sciences
Project
https://sites.psu.edu/mcreu/files/formidable/2/2024-07-23/APetrov-2024-MC-REU-Poster-RC1.pdf
Project Video
Project Abstract
Cannabis sativa L. (C. sativa), also known as industrial hemp, is a dioecious crop closely related
to marijuana (Andre et al., 2016; Crini et al., 2020; Mistry et al., 2021; Patel et al., 2021).
Industrial hemp belongs to the Cannabis genus and has less than 0.3% of psychoactive
tetrahydrocannabinol (THC) and higher levels of cannabidiol (CBD) whereas marijuana has more
than 0.3% of THC (Andre et al., 2016; Mistry et al., 2021; Patel et al., 2021). Hemp is used in
various industries including but not limited to the textiles and paper industry, insulation and
building materials, horticulture, animal nutrition, food and beverages, nutraceuticals,
cosmetics, hygiene, agrochemistry, and energy production (Andre et al., 2016; Crini et al.,
2020). This study aims to maximize the positive yields of industrial hemp through the utilization of AI and Machine Learning methods. Industrial hemp seeds were plated on Petri dishes with Agarose Media and placed under arrays of cameras. Multiple techniques for improving image readability were employed, such as minimizing condensation on the lid of the Petri dishes, preventing glare, and using colored fabric as a background. An AI model that consisted of Convolutional Neural Networks (CNNs) and binary classifiers was trained on the image dataset and is currently converging to a loss of 0.58 and a test accuracy of 70%. These statistics are expected to be misleading due to a very small dataset size, but the initial results show potential for improvement by adding new batches of growth data. Subsequent experiments will explore monitoring other growth factors and a larger variety of seeds.
Evaluate this Project
Use this form link to provide feedback to the presenters, and add your project evaluation for award(s) consideration.