My professor asked for some images of cool stuff I worked on for a presentation he is giving. So here is me moving fast enough to cause motion blur :) The code is using the camera to detect people, and then project the bounding box down in the lidar frame, and mask all the lidar points within that cone.
Anyway, if someone is familiar with super fast agglomerative clustering code in C++ (or even python, if it's efficient), please share it with me!

  • 8
    oh wow DevRant butchered the quality of the image.....
  • 4
    I don't know what those green and red lines are but they do look professional and impressive
  • 5
    @LotsOfCaffeine The points are the 2D Lidar beams hitting obstacles. Green ones are valid and will be used for localization, red ones correspond to people so I am masking them in the algorithm.
  • 5
    @LotsOfCaffeine In the original image you can actually see the feet pattern (half-circles).
  • 4
    I knew it was you before I opened the rant. 😆
    This is good stuff.
  • 2
    Is that a view from the top ? Are you projecting the 3D points to the ground plane ? Or is it a cross-section (more like a really thin cuboidal slab to contain the points) ?
  • 3
    @-red It's a 2D lidar so I only have 1 horizontal "slice". The only information I have is the distance and heading from the lidar, so it's actually more like 2D points, and we assign then a given height assuming the lidar beams are parallel to the floor. Yes, it's a top-view of the scene
Add Comment