Computational Imaging and Illumination for 3D Acquisition: Research at the NU Comp Photo Lab

Date

July 10, 2015

Speaker

Oliver Cossairt

Affiliation

Northwestern University

Overview

Computational imaging and illumination plays a central role in many modern three-dimensional imaging techniques. In this talk, I will provide an overview of several 3D imaging technologies pioneered by the NU Comp Photo Lab, highlighting 3 main research projects. First, I will introduce a novel structured light technique called Motion Contrast 3D scanning (MC3D) that maximizes bandwidth and light source power to avoid performance trade-offs in structured light 3D acquisition. The technique allows 3D laser scanning resolution with single-shot speed, even in the presence of strong ambient illumination, significant inter-reflections, and highly reflective surfaces. Next, I will present research on Incoherent Holography, which enables 3D digital refocusing by engineering the camera’s Point Spread Function (PSF). The technique can be used to capture a hologram without illuminating the scene with a coherent laser, making it possible to acquire holograms even for passively illuminated scenes. Finally, I will introduce our work using photometric stereo to measure the surface shape of several of Paul Gauguin’s prints and drawings housed at the Art Institute of Chicago. In this work we characterize surface topology to better understand the artists production methods, helping to resolve longstanding art historical questions about the evolution of Gauguin’s printing techniques.

Speakers

Oliver Cossairt

Oliver Cossairt currently holds the Lisa Wissner-Slivka and Benjamin Slivka Junior Chair in Computer Science at Northwestern University. Before joining Northwestern, he developed several key advances in 3D Display technology while earning my Masters at the MIT Media Lab and as an Optical Engineer at Actuality Systems, resulting in 8 patents. He earned his Ph.D. from Columbia University, where he was awarded an NSF Graduate Research Fellowship for his thesis work on the theoretical limits of computational imaging. Prof. Cossairt is currently director of the Computational Photography Laboratory at Northwestern University (http://compphotolab.northwestern.edu), whose research consists of a diverse portfolio, ranging in topics from optics/photonics, computer graphics, computer vision, and image processing. He has written a number of high-impact publications including ACM SIGGRAPH, IEEE Pattern Analysis and Machine Intelligence, and IEEE Transaction on Image Processing, and received Best Paper (2011) and Honorable Mention (2014) awards for publications at the IEEE International Conference on Computational Photography. He has given a keynote talk at the IEEE Computational Cameras and Displays Workshop (2013), as well as invited talks at a number of conferences including: AAAS (2015), COSI (2013,2015), CLEO Europe (2015), ICIP (2015), IPAM Computational Photography (2015), and FRINGE (2013). He has co-organized several workshops, serves on 5 conference program committees annually, and as Associate Editor for the IEEE Transactions on Computational Imaging. His research projects have garnered funding from numerous corporate sponsorships (Google, Rambus, Samsung, Omron) and federal funding agencies (ONR, NIH, DOE, NSF CAREER Award).