Object placement using Point Cloud Library (PCL)

PROJECT

This project focuses on identifying possible drop points for a robot to place an object using a point cloud obtained from an RGB-D sensor. The sensor, mounted on the robot (Lucy), captures color and depth information, which is then processed to determine suitable locations for object placement.

Dependencies

  • Python 3
  • NumPy
  • Matplotlib
  • SciPy
  • Point Cloud Library (PCL)

Methodology

  • The point cloud (cloud.pcd) is transformed based on the camera’s pose relative to the robot’s base.
  • Points are then ranked using a point ranking algorithm, considering their angles with respect to the Z-axis.
  • The output is visualized in a 3D plot, where suitable drop points for object placement are highlighted in red.

Github repository: The repository can be found here