Could 'Pseudo-LiDAR' Accelerate Self-Driving Cars?

National Science Foundation-funded researchers from Cornell University  have found a simpler, less expensive alternative to the LiDAR sensors currently used in autonomous cars to detect objects. The Cornell team’s new method, "pseudo-LiDAR," uses two inexpensive stereo cameras to detect objects with near-LiDAR accuracy. The researchers found that analyzing the captured images from a bird’s-eye view, as opposed to the more traditional frontal view, more than tripled their accuracy – making stereo cameras a viable and low-cost alternative to LiDAR. Ultimately, the researchers see these cameras being used as the primary way of identifying objects in lower-cost cars, or potentially as a backup method in higher-end cars that are also equipped with LiDAR.