A study on ego-motion estimation based on stereo camera sensor and 2G1Y inertial sensor with considering vehicle dynamics

The vision sensor on a vehicle platform with dynamic motion is affected severely by disturbances that occur owing to vehicle dynamic movement, which results in the degradation of outputs from vision sensor such like the distance between ego-vehicle to objects. To get around this phenomenon, it is necessary to estimate vehicle ego-motion, that is, tri-axis linear velocities and angular velocities that can be utilized to make up for the contaminated outputs from vision sensor. This study proposes a novel method for vehicle ego-motion estimation by using the optical and disparity flows from vision sensor and 2G1Y (longitudinal/lateral acceleration and yaw-rate) measurements from the inertial sensor in vehicle. The approach considers vehicle platform characteristics to estimate ego-motion more accurately. To the end, vehicle kinetics are augmented into the evolving equation represented in a form of state space equation for applying Kalman filter and the vanishing point is to be estimated online to enhance the information of optical flows because the optical flow radiates from a vanishing point that might vary owing to vehicle dynamic motion. To show the effectiveness of the proposed algorithm, experiments have been carried out on real vehicle environments and the results confirm its effectiveness and accuracy.

Language

  • English

Media Info

Subject/Index Terms

Filing Info

  • Accession Number: 01711721
  • Record Type: Publication
  • Files: TRIS
  • Created Date: Jun 19 2019 3:05PM