The project aims to evaluate the feasibility of reconstructing 3D point cloud of plants from 2D images from different angles, and estimate leaf angles from the 3D point cloud.

  • Work with robotics team to develop an imaging prototype that can automatically capture 2D images of plants from various angles
  • Apply StereoVision and Structure-From-Motion algorithms to reconstruct 3D point clouds of plants using Python and OpenSFM
  • Segment individual leaves and estimate leaf angles using Python and OpenCV