Computer Vision Solution


Computer Vision Solution for Each Level of Autonomous Vehicles

Phantom AI’s computer vision technology provides fundamental building blocks for autonomous vehicles. Armed with a Deep-learning based detection front-end and geometry-based filtering back-ends, PhantomVisionTM can provide not only automotive reliable but also customizable and extensible vision solutions to Tier 1s and OEMs.

Computer Vision Solution

Level 2 Autonomous Vehicles


Phantom AI has developed a robust level 2 solution with automotive reliability designed in from the beginning. Starting with comprehensive system requirements, each fault-tolerant subsystem is constantly validated through the use of simulation and real-world testing.

PhantomVision™: Deep-learning based computer vision solution

  •   US/Euro NCAP capable perception using three forward-facing cameras
  •   Support for vehicle, pedestrian, bicyclist, free-space, traffic sign and traffic light detection

PhantomFusion™: Modular sensor fusion solution

  •    Traffic environment recognition and prediction
  •    Camera and Radar based L2 solution, but configurable for a mixture of sensor types

PhantomDrive™: Vehicle control solution

  •    Adaptive Cruise Control with cut-in vehicle handling
  •    Automatic Lane Keeping and Automatic Lane Change
  •    Automatic Emergency Braking : highway and city (pedestrians/bicyclists)

Level 4 Autonomous Vehicles


Phantom AI sees a future in mobility-as-a-service. By extending the features and capabilities already present in the Level 2 software stack, Phantom AI is developing an optimal platform and sensor suite for level 4 functionality.