15 to 30 of September

Year 1

Posted by Daniela Rato on September 30, 2021

Calibration with real hardware

This week's progress:

  • Placement of the sensors mimicking the simulation setup
  • Recording of rosbag using real data
  • Adaptation of methodology for a centered camera
  • Setting initial estimate in real setup
  • Collecting data and obtaining collections with real setup
  • Callibration with data obtained with real hardware

TODO

  • Check inconsistencies in calibration results with real data

Tf tree in simulation vs tf tree with real hardware

Simulation

(zoom)

Real setup

(zoom)


Demonstration of initial estimate


Demonstration of data collecting


Demonstration of calibration procedure (initial guess noise 0, 0 - tol 10e-6)


Demonstration of calibration procedure (initial guess noise 0.2, 0.2 - tol 10e-6)

Issues

calibration in simulation mode - open

calibration with real hardware - open

Change resolution of astra camera to SXGA or XGA - open