GOALS

• Convergence
  Convergence was one of our main goals. Since the previous team had struggled to get much convergence with their particles, we decided that a large part of our time would be dedicated to fixing this. (Without this convergence the MCL algorithm can not work!)


• Panorama Stitching
  Stitching the panorama correctly, so it contained all the rotation images and was not corrupted in any way. This was also a large challenge for the previous group; if the panorama wasn't in a good state this could really affect the algorithm as the robot wouldn't be able to accuratly judge where it was.


• Finding where home is in the Panorama
  Another goal for us was to find where the home image (rotation 0) was within the panorama. Without this the robot wouldn't know where to return to, and would just randomly turn.


• Image comparison
  Our final goal was to change how the images were being compared. With this we could make an accurate estimation of where the robot is currently looking, and allowed us to scale the weights for each particle.