John Deere's autonomous tractor classifies each camera pixel in about 100 milliseconds

The fully autonomous tractor John Deere revealed at CES in January 2022 uses six pairs of stereo cameras for 360-degree obstacle detection. Deere says the images those cameras capture “are passed through a deep neural network that classifies each pixel in approximately 100 milliseconds,” and the result determines whether the machine continues moving or stops because it has detected an obstacle. The tractor also checks its position against a geofence to within less than an inch of accuracy.