Antsy 3D

Lightweight point cloud annotation tool for instance-level segmentation using fat markers and raytracing.

Notes:

• This has now evolved into a more sophisticated LiDaR annotator: LiDaR Annotator by B. Wang et. al. I recommend using that instead.
• Sorry, current version using skinny (1-pixel) marker. :( Another option is to use Three.js's built-in octotree and color entire boxes.

Usage

No installation needed! This is an in-browser interactive view of your models.

1. To run, double-click on index.html to launch the app.
2. To label, simply hold your mouse down and drag across the screen.
3. To export your labelled points, click on the "save" button to the top-left.

Need to rotate or move your point cloud? To change between move mode and label mode, click on the respective buttons in the top-left:

• In move mode, dragging will reorient the point cloud. Scrolling with zoom in and out.
• In label mode, dragging will added red dots, representing labelled points.

Point Cloud Format

js/output.js defines a global variable data, a dictionary mapping unique point cloud names to point cloud data; point clouds have a vertices property, a list of vertices, each with .x, .y and .z properties.

e.g., To access the x position of the first point in the point cloud named frog, you would access data.frog.vertices[0].x. Simple example of output.js:

var data = {
   'frog': {
      'vertices': [
          {'x': 3.5, 'y': 3.5, 'z': 5.6},
          {'x': 1.0, 'y': 1.2, 'z': 3.4}
      ]
   }
}