Report an accessibility problem

Engineering | Undergraduate Research

Change detection and analysis in manufacturing video data

Change detection and analysis in manufacturing video data

In advanced manufacturing applications, inline imaging sensors have been widely adopted for capturing real-time information about the product and manufacturing process. The acquired data are often real-time videos of product(s). They are vital resources for detecting the occurrence of defects, process instability, and system faults. The data form as videos, however, has posed challenges for data processing, modeling, and analysis. First, raw videos require cleaning to remove noise and enhance image quality. Second, each video may have useful frames and uninformative frames. How to find the useful information out of the mixed images remains a problem. Third, consecutive frames in a video may contain temporal connections (e.g., correlation), which reflect the build or formulation process of a product and worth further exploration. In this research project, the student will perform video processing, modeling, and analysis to detect interesting changes in a video and explore the temporal connections across frames. The objectives will be providing the student knowledge in handling imaging data, developing new thoughts for video processing, and generate new knowledge for data-driven decision-making in manufacturing. The student will be given the video data and use programming software to develop novel methods and validate them on the data.

Courses: statistics, mathematics

Skills: programming (R/Python/Matlab)

Spring 2022, ~3-5 hours/week, volunteer

Please send email to the following address: and include your resume and transcripts, as well as a brief introduction to your background, research interest, and objectives for participating in this research project.


, ,

Posted on