Researchers Enable Today’s Vehicle and Smartphone Cameras to See Around Corners
Algolux and University Scientists Develop Non-Line-of-Sight Technology That Photographs Hidden Objects Using Conventional Camera Sensors
Conference on Computer Vision and Pattern Recognition (CVPR), Long Beach, California – June 18, 2019 – Algolux has announced that a team of researchers from Algolux, the University of Montreal, and Princeton University has developed a new method that lets conventional color cameras — the ones in your smartphone or in a vehicle camera — see hidden objects that are occluded by walls or other scene objects.
The team achieved unprecedented resolution for non-line-of-sight (NLOS) imaging by being able to see objects in high-resolution and color around corners for the first time. The researchers from academia and industry were able to reconstruct high-quality images of traffic signs (see Figure 1) and other 3D objects without looking directly at those objects.
Figure 1. Reconstruction of traffic signs with high-resolution color Non-Line-Of-Sight imaging using conventional CMOS cameras sensors.
The authors were able to accomplish these results with existing conventional CMOS camera sensors and a change in illumination method, needing only a small change to a car’s headlights or a smartphone’s flash. This research breakthrough opens a path to practical implementation across numerous markets.
Specifically, Algolux believes this technology can further strengthen the ability for autonomous vehicles to safely navigate in difficult road scenarios even when the view is blocked by obstructions or vehicles, deliver increased security for video surveillance, and enable additional use cases for smartphones, augmented reality, and medical imaging (Figure 2).
Figure 2. Example of applying high-resolution Non-Line-Of-Sight imaging for a real-world driving application.
The findings will be published at the prestigious CVPR computer vision conference this week and were chosen to be presented as one of the select few oral presentations.
More details on this research can be found on the dedicated project page at https://www.cs.princeton.edu/~fheide/steadystatenlos.
For information about Algolux’s research projects, please visit https://algolux.com/research
Algolux provides machine learning optimization and embedded software for robust autonomous vision – empowering cameras to see more clearly and perceive what cannot be sensed with today’s imaging and vision systems. Computer vision is at the heart of autonomous cars, ADAS, mobile devices, security cameras (IoT), AR/VR, robotics, drones, and medical equipment, leading the next wave of market growth and social impact. Developed by an industry-recognized team of researchers, our award-winning machine learning technologies and end-to-end methodology address the complex task of developing optimal imaging and vision systems. This approach improves vehicle safety and system performance while reducing the program risks faced by product development teams, such as cost, time-to-market, and scalability.
Learn more by visiting us at https://algolux.com and join the discussion on our LinkedIn and Twitter.
© Copyright 2019 Algolux, Inc. Algolux, the Algolux logo, CRISP, CRISP-ML, CANA, and other designated brands included herein are trademarks of Algolux in the United States and other countries. All other trademarks are the property of their respective owners.
+1 (877) 424-9107