Vision playground is as the name suggest just a playground project for 2D/3D test and simulations in python.
Core features in vision playground
- Camera and/or Lidar simulation
- 3D reconstruction
- 2D detection
Give a 3D scene and a sequence of poses and it'll save all the images of the 3D scene at those poses.
Give a 3D scene and a sequence of poses and it'll get the RGBD images of the 3D scene at those poses and do a 3D reconstruction from them.
Give an image and it'll run ML object detection on it.
- Feature detection and matching
- 3D reconstruction from RGB
- 2D segmentation