Optical Flow for Bridge Component Annotation
Department
Computer Science
Major
Computer Science
Research Advisor
Chen, Genda
Advisor's Department
Civil, Architectural and Environmental Engineering
Funding Source
Grant
Abstract
Optical flow is an algorithm that follows apparent motion of objects in a video. It could be used to track bridge structures such as: piers, girders, cross-girders, and more. MODOT employees connect to our website, upload a video of a bridge they have taken themselves, preferably with a drone, and annotate structures themselves with an integrated VGG annotator. Optical flow expedites the process by automatically labeling structures between annotated frames. In the end, the employee saves time, and we have more training data.
Biography
Rowan is a senior at the Missouri University of Science and Technology studying Computer Science. He is on track to graduate in Fall 2024 with his bachelor's degree. Ever since Rowan was in elementary school, he was setting up Minecraft servers. Now, he built his own server to self-host web services and train neural networks. Along with his personal projects, in his professional life, he has taken opportunities to work in data science.
Research Category
Engineering
Presentation Type
Oral Presentation
Document Type
Presentation
Location
Havener Center - Carver Room
Presentation Date
10 April 2024, 9:00 am - 12:00 pm
Optical Flow for Bridge Component Annotation
Havener Center - Carver Room
Optical flow is an algorithm that follows apparent motion of objects in a video. It could be used to track bridge structures such as: piers, girders, cross-girders, and more. MODOT employees connect to our website, upload a video of a bridge they have taken themselves, preferably with a drone, and annotate structures themselves with an integrated VGG annotator. Optical flow expedites the process by automatically labeling structures between annotated frames. In the end, the employee saves time, and we have more training data.
