Machine Vision - Forward Looking Cameras


Function/ SW-Module

Soil Detection

Description
  • Detection of obstructions (soil, dirt, mud etc.)
  • Image is separated into tiles
  • Sensitivity is reduced using temporal filtering
  • Light-weight for embedded system
  • Also suitable for rear-/surroundview cameras

Function/ SW-Module

Online Calibration

Description
  • Highly accurate estimation of camera extrinsics (position, orientation)
  • Computed during operation and running in background
  • Compensates long term changes and displacements of camera position
  • Also suitable for rear-/surroundview cameras

Function/ SW-Module

Traffic Sign Recognition

Description
  • Fast radial symmetry transform for fast circular object detection
  • Sign classification with AdaBoost/SVM (one against all algorithm)
  • Haar feature based approach

Function/ SW-Module

Vehicle Recognition

Description
  • Intelligent context dependent search strategy
  • Classification into passenger cars and trucks
  • Pitch angle invariant distance estimation
  • Scale change based velocity estimation

Function/ SW-Module

High Beam Assist

Description
  • Support vector machine based classifier
  • Intensity over time (reflector detection)
  • LED taillight detection
  • Guard rail detection
  • City detection (50Hz streetlamp detection)

Function/ SW-Module

Lane Recognition

Description
  • Multi model tracking of line markings
  • High level scene analysis
  • Optical flow based pitch rate estimation
  • Full automatic camera angle calibration
  • Minimalistic feed back loop between tracking and image processing

Function/ SW-Module

Stereo Obstacle Detection

Description
  • Detection of obstacles (no classification)
  • Ground plane estimation
  • Tracking of objects
  • 3D coordinate calculation of obstacles
  • Calculation of relative speed of the obstacles

Machine Vision - Surround View Camera


Function/ SW-Module

Structure from Motion/ Motion Stereo

Description
  • 2D Feature point detection and tracking
  • Ground plane estimation
  • Camera based ego motion estimation
  • 3D reconstruction (triangulation)

Function/ SW-Module

Backover Protection

Description
  • Intelligent clustering of 3D points
  • Detection of obstacles
  • Calculation of 3D coordinates of obstacles

Function/ SW-Module

Crossing Traffic

Description
  • Detection of approaching vehicles (passenger cars, trucks, bicycles, motorcycles)
  • Calculation of “time-to-collision”

Function/ SW-Module

Rear End Collision Warning

Description
  • Forward looking functionalities, lane recognition and vehicle recognition adapted to megapixel camera and fish eye optics
  • Classification of approaching cars
  • Calculation of time-to-collision

Machine Vision - CMS


Function/ SW-Module

Detection of approaching objects during stand-still of ego-vehicle

Description
  • Optical flow based approach
  • Detection of approaching vehicles (passenger cars, trucks, bicycles, motorcycles)
  • Calculation of “time-to-collision”
  • Possible application e.g. „Open Door Warner“, „Safe Exit“ etc.

Function/ SW-Module

Vehicle recognition

Description
  • Based on HoG and SVM approach
  • Intelligent context dependent search strategy
  • Classification into passenger cars and trucks
  • Scale change based velocity estimation
  • Possible application e.g. LCA (lane change assist)

Function/ SW-Module

Overlays

Description
  • Giving support to the driver during certain maneuvers

Machine Vision - Commercial Vehicles


Function/ SW-Module

Stereo Obstacle Detection

Description
  • Detection of obstacles (no classification)
  • Ground plane estimation
  • Tracking of objects
  • 3D coordinate calculation of obstacles
  • Calculation of relative speed of the obstacles
  • Trajectory calculation
  • Possible application: „Turning Assist“ for trucks

+49 (0)8382 9114 701