Avonelle working in the lab next to a UAV
Author: Avonelle Lindon | Major: Biological Engineering | Semester: Spring 2024
Precision Agriculture using Computer Vision
I am studying Biological Engineering with a minor in sustainability and received an Honors College Grant the spring semester of my sophomore year. I have been working with Dr. Koparan to quantify crop coverage and weed pressure using computer vision.
The main objective of this research is to create an efficient and accurate method for quantifying weed presence in agricultural fields. By using UAV (Unmanned Aerial Vehicle) technology equipped with advanced cameras, we can capture high-resolution images of soybean fields. These images are then processed using computer vision algorithms to analyze the Normalized Difference Vegetation Index (NDVI), which helps identify vegetation in images.
This model aims to make herbicide application more efficient, which has significant impacts for sustainable agricultural practices. Traditional weed management often uses uniform herbicide spraying, leading to overuse and increased environmental impact. The goal of the computer vision model is to allow for precise, real-time weed management by identifying specific areas with high weed pressure. This targeted approach can significantly reduce herbicide use, minimizing its harmful effects on the environment, such as soil degradation and water contamination.
How did you choose your topic? How did you find your mentor?
My fascination with sustainability and the integration of technology in agriculture led me to explore precision agriculture. After emailing several professors whose work interested me, I met with Dr. Cengiz Koparan and found his research on computer vision in agriculture particularly compelling. After deciding to work in his lab, I mentioned that I was interested in applying for the Honors College Grant, and he helped me with the application process, providing guidance and support.
What did you learn?
Throughout this research, I gained in-depth knowledge about computer vision technology and its applications in agriculture. I learned how to develop and test models that can accurately analyze images to determine weed density. Additionally, I enhanced my understanding of statistical analysis, scientific research methods, and scientific writing. This experience has also honed my problem-solving skills and my ability to conduct independent research.
What challenges did you face in your research? How did you overcome them?
One of the major focuses of this semester was testing the accuracy and functionality of the computer vision model compared to traditional methods like ImageJ analysis. This required extensive analysis of field images. I had to process a large number of images and create a program to analyze them. I ran into several issues, particularly with coding and data processing. However, I collaborated with another undergraduate student who had more experience in coding, and was able to troubleshoot the issues that came up. This collaboration also provided an opportunity to learn from my peers and improve my programming skills.
What role did your faculty mentor play?
Dr. Koparan played a crucial role in my research journey. During our weekly meetings, Dr. Koparan answered any questions and helped me troubleshoot issues I encountered. He also helped me refine my research questions and develop my methodology.
What’s next?
Over the summer, more data will be collected from UAV flights. Next semester, I will continue my research by conducting further data analysis, testing of the model and preparing for presenting my research. I hope to attend a conference to present what I have worked on, sharing my findings with the scientific community. After college, I aim to work as an engineer, focusing on implementing sustainable solutions in the food and agriculture sectors. By leveraging technology to enhance sustainability, I hope to contribute to creating more efficient and economically viable solutions.