NODAR, the leading provider of next-generation stereo vision technology for autonomous vehicles, announced today that it has been awarded two additional U.S. patents for its non-rigid stereo vision system, which uses computer vision to automatically calibrate multiple cameras and produce the high fidelity 3D information required for self-driving vehicles to navigate.
Mounted independently, NODAR's long-range, high-resolution, camera-based sensors provide 360-degree views of an AV's environment in real time, significantly improving the safety of autonomous vehicles and advanced driver-assist systems.
NODAR has been awarded U.S. Patent Nos. US11321875B2 and US11321876B2 for a stereo vision system that allows for relative motion between two or more cameras without degrading the estimates indicated on a depth map. The system tracks camera parameters and synchronizes all cameras to rectify images even as camera positions shift when cars are in motion. Computer-implemented calibration eliminates the need to calibrate the system during production, or manually during routine operation.
NODAR was awarded a related patent (U.S. Patent No. 17/365,623) for its non-rigid stereo vision camera system earlier this year.
"These patents are a further confirmation that our camera-based 3D vision technology is uniquely effective, promising an entirely new level of safety for fully autonomous vehicles and driver-assist systems," NODAR founder and chief executive Leaf Jiang says. "Using data from multiple cameras allows the system to deliver highly accurate ultra-long-range sensing, detecting even very small objects in the road and their distance from the car. It's a huge advance in the safe operation of AVs and ADAS, particularly at highway speeds, with the added advantage of a competitive price point."
The company's flagship product, Hammerhead™, automatically maintains alignment between independently-mounted cameras using patented computer vision algorithms, advanced processing, and camera technology. The sensor system ensures reliable depth measurements despite any road and vehicle vibrations, and also makes it possible to mount cameras far apart, creating the wide-baseline triangulation that optimizes the system's range and accuracy.