Multi-Camera and Multi-Algorithm Architecture for Visual Perception Onboard the ATLASCAR2.

My Master's Thesis Blog

Visual Perception of the road lines/boundaries

Visual Perception of the road lanes/boundaries includes to implement artificial vision techniques in order to extract the road lines in an image that came from a camera.

My purpose with this thesis is to develop a solution that complements a great ROS software architecture with Visual Perception of road lanes algorithms.

Week 0 | (13/02 - 15/02)

  • ROS Installation and tutorials.


Week 1-2 | (18/02 - 01/03)

  • ROS Tutorials (continuation).
  • ROS Workshop.
  • Preliminary Report.
  • GitHub repository creation.
  • Website creation.


Week 3-4 | (04/03 - 13/03)

  • First Lar Meeting presentation.
  • Model the physical structure needed to fix the camera.
  • Definition of the software architecture.
  • First contact with the camera.
  • Intrinsic Calibration


Week 5 | (14/03 - 21/03)

  • Start to assemble the camera support on car.
  • Test and study some algorithms to implement on the developed architecture.
  • Resolution of several incompatibilities problems in the algorithms.


Week 6 | (21/03 - 28/03)

  • Start of installation latest Ubuntu and ROS versions on ALTASCAR's computer.
  • Finalization of the study of the first implemented algorithm.


Week 7 | (28/03 - 04/04)

  • Brainstorm about referee node.
  • Second algorithm simulation.
  • Starting of the implementation of the referee node.


Week 8 | (04/04 - 11/04)

  • Referee package implementation.
  • Atlascar 2 inspection.


Week 9-10 | (11/04 - 25/04)

  • Advanced algorithm ROS package creation.
  • Performance architecture study.
  • Initial images resize.
  • Study of how to synchronize two topics.
  • Starting of dissertation writing.


Week 11 | (02/05)

  • Prepation of the camera setup on ATLASCAR.
  • Second camera issue.
  • Continuation of dissertation writing.

Week 12 | (09/05)

  • Boxes assemble.
  • Camera software installation on ATLASCAR computer.
  • Installation of my package on ATLASCAR computer.
  • Continuation of dissertation writing.

Week 14 | (23/05)

  • GPS trobleshooting.
  • Experimental data aquisition.
  • Parametrisation of the advanced algorithm.
  • Continuation of dissertation writing.

Week 15-16 | (08/06)

  • Deep Learning Approaches

Week 16-17 | (20/06)

  • Tests and experiments.