Professor Goyal’s Record-Breaking Long Distance NLOS Imaging Featured in PNAS

by Caroline Amato

Feihu Xu / USTC

The March 9, 2021 issue of Proceedings of the National Academy of Sciences of the USA (PNAS) published findings from a collaboration between BU ECE Professor Vivek Goyal, and researchers from the University of Science and Technology of China. These findings result from a demonstration of non-line-of-sight (NLOS) imaging and tracking from a distance of 1.43 kilometers; this is about three orders of magnitude farther than reported to date in NLOS trials.

NLOS imaging is a method for recovering parts of hidden scenes from indirect light paths that scatter several times between the source of illumination and the detector. This advanced imaging can be applied in fields such as robotics, navigation, scientific imaging, and surveillance.

Despite recent advances in the technique, NLOS imaging has yet to gain widespread popularity, in part because prior demonstrations have been constrained to a short range. This limitation is imposed by very weak signals that survive multiple diffuse reflections. Both experimental and algorithmic innovations in Professor Goyal’s approach contributed to the significant increase in range in the newly-published demonstration, compared to previous experiments.

At 1.43 km, about one informative photon is detected per seven quadrillion emitted photons. A high-efficiency, low-noise dual-telescope system developed at the Chinese university combined with advanced reconstruction algorithms were able to work together to enable this record-breaking demonstration of NLOS imaging.

Professor Goyal’s many accolades include the IEEE Signal Processing Society Best Paper Awards in both 2017 and 2019, as well as the IEEE International Conference on Image Processing Best Paper Award in 2014. He is a fellow of IEEE and OSA. He won the 2013 MIT $100K Entrepreneurship Competition Grand Prize, and his PhD student, Joshua Rapp, received the IEEE Signal Processing Society Young Author Best Paper Award in 2020 for work he co-authored. Professor Goyal continues to pursue innovation in his work with computational imaging with weak and unusual signals.