Indoor localization system based on a gyro and visual passive markers

From AIRWiki

Jump to: navigation, search
Indoor localization system based on a gyro and visual passive markers
Short Description: This project aims at developing autonomous moving systems based on a gyro and passive markers
Coordinator:
Tutor: MatteoMatteucci (matteo.matteucci@polimi.it), SimoneCeriani (ceriani@elet.polimi.it), DavideMigliore (d.migliore@evidence.eu.com)
Collaborator:
Students: DarioCecchetto (dario.cecchetto@mail.polimi.it), LorenzoConsolaro ()
Research Topic:
Start: 2009/06/01
End: 2011/06/01
Status: Closed
Level: Bs
Type: Thesis

Contents

What we use

  • Eclipse for C/C++
  • OpenOCD
  • An STR9 Cam based RVS module from ST
  • A STM32 Gyroscope based RVS module from ST
  • Kalman algorithm adapted for 2D.
  • Fast algorithm for Landmarks recognition

The Project in short

Using SLAM technique we want to develop an indoor system based on a gyroscope, an accelerometer and a camera.


SLAM Algorithm we use

  1. . Predict new robot pose with Gyroscope
  2. . Modify the covariance
  3. . For each of the M observed landmarks i:
    1. . Compute hi and its Jacobian.
    2. . For each dimension j of the observation hi:
      1. Compute the scalar Sij
      2. Computation of Kij.
      3. Update the filter state vector
      4. Update the filter covariance Pk | k
  4. . If necessary, introduce new landmarks in the map.

References