The PerCV™ System provides a complete scene understanding to improve autonomous mobility. It accepts input from a variety of sensors, including cameras, LIDAR and radar, and uses deep learning pipelines to process the data. PerCV identifies and classifies objects from the sensor data and indicates their locations with respect to the platform. The deep learning algorithms are trained to identify a variety of objects, including:
- Traffic sign classification
- Traffic signal classification
- Lane detection
- Vehicle detection
- Pedestrian detection
PerCV can improve ground navigation by providing critical information to both drivers and autonomous systems alike. Advanced driver-assistance system (ADAS) functionalities can be implemented, reducing the risk of accidents due to human error.