LeddarVision perception software provides highly efficient detection, classification and tracking of objects for vehicle automation up to Level 2+ based on camera and radar sensor data. LeddarVision solutions meet the perception requirements of L2 and L2+ driver-assist applications without the need to rely on LiDAR sensors.

L2/L2+ Environmental Model

    • The video sequence below presents the LeddarVision perception in action based on a dataset from nuScenes captured with a system using 6 cameras and 5 radars.

(LiDAR data is being used for the sole purpose of visualization).

 

Key Features for L2-L2+

  • Level 3 performance at Level 2 cost
  • Real-time detection, classification, tracking
  • Based on camera and radar; LiDAR data not necessary
  • 3D bounding boxes
  • Object position, orientation and velocity
  • Support for multiple sensors

LeddarVision Environmental Model

LeddarVision perception solutions provide a complete environmental model and include AI and DNN processing to support the highest levels of autonomy.

  • Detection
  • Calibration
  • Segmentation
  • Identification & tracking
  • Localization & mapping
  • Lane segmentation
Back to LeddarVision Overview