Mechanical and Civil Engineering Seminar
Smaller Than the Eye Can See: Vibration Analysis with Video Cameras
Justin Chen is currently a Technical Staff member at MIT Lincoln Laboratory in the Advanced Capabilities and Technologies group. His main research interests are the development of novel sensor systems and methodologies to improve infrastructure resilience through quantitative condition assessment. He received his Ph.D from MIT in 2016 in Structures and Materials working in the Laboratory for Infrastructure Science and Sustainability, headed by Prof. Oral Buyukozturk. Before graduate school, he was an assistant staff member at MIT Lincoln Laboratory and he received a B.S. in Physics from the California Institute of Technology in 2009.
The tools that we use to assess the condition of infrastructure and conduct inspections have evolved over the years. Originally, inspectors used their senses of sight, hearing, and touch to evaluate structures, but now they use advanced equipment such as ultrasound and ground-penetrating radar which can image through materials and make finer measurements beyond human capabilities. Recent developments in computer vision have created a "motion microscope", where small imperceptible motions in videos can be amplified and easily seen and enables quantitative analysis of motions from video as a basis for vibrational analysis of objects. This technique and related work is collectively called motion magnification.
This talk will present the application and development of motion magnification for characterizing the motions and vibrations of machinery and structures towards the goal of resilient infrastructure. I will describe the methodology and theory behind motion magnification, the workflow from video to modal information, and methods for processing the numerous signals collected from video data. I will show experimental efforts in using the technique ranging from video measurements of small structures and rotating machinery to field measurements of a vertical-lift bridge. Future prospects for camera based measurements for a variety of applications in infrastructure resilience spanning many length scales will be discussed.