Gabriel St.Jean
UGA Senior Capstone Design: Spray Quality Monitor for Informed Pesticide Decisions
PROJECT DESCRIPTION
The Spray Quality Monitor for Informed Pesticide Applications is a UGA Capstone research & development project sponsored by UGA Tifton and kick-started in the 2021 - 2022 academic school year. This project strives to obtain the spray quality of a pesticide spray in real-time by the use of computer vision and a camera module to measure spray droplets. In the application, the camera module will be mounted in front of a boom sprayer so that images of spray droplets can be captured from the applicator. A microcontroller will then be connected to the camera module so that spray droplets can be extrapolated from the image. Visual feedback to the user will exist in the form of a pie chart where changes in spray quality can be viewed in real-time. A simulation of the Spray Quality Monitor in an infield setting has been created by the team with the use of pre-captured images obtained from the team's test stand. This can be seen in the video below.
​
​
Programming languages and software: Python and OpenCV
* Please contact me at my email to see the private code repository
​
​
​
​
2022 UGA Capstone Showcase Award Winner For Innovative Computerized or Software Design
As a result of my team's presentation performance at the 2022 UGA Capstone Design Showcase and their excellence in the documentation and prototype-development phase, I am proud to announce that the Spray Quality Monitor won the award for the best Innovative Computerized design at the Showcase. I would like to give a special thanks to the teams' mentors Dr. Mark Haideckker, Dr. Rawad Saleh, and Dr. Roger Hilten whose impact not only contributed to the team satisfying the needs of their sponsor and client but also led us towards winning the award at the showcase.
Contributions
-
Utilized Canny edge detection to identify spray droplets within an image and each frame of a video.
-
Created a trackbar GUI application to manually fine tune the parameters of the Canny edge detection algorithm.
-
Developed a measurement algorithm to convert the pixel area of spray droplets into a micron measurement.​
-
Created a pie chart that executed alongside the spray droplet detection algorithm to visualize real-time changes in spray quality of the application.
-
Consulted with the project's client on a bi-weekly basis to discuss project requirements and to provide progress reports in the form of a presentation.
Reflection
The Spray Quality Monitor was a success in satisfying the expectations of the client, sponsor, and the university because a program was created that could identify spray droplets from an image and visualize the micron-measurement of each droplet to the user. Improvements to the accuracy and the noise filtering capabilities of the spray droplet detection algorithm are to be desired. Nonetheless, the team's progress in the spray droplet identification program and prototype development phase has provided sufficient grounds for the project to be continued by a new capstone team for the 2022 - 2023 academic school year.
