mithi / point-cloud-filter

Scripts showcasing filtering techniques applied to point cloud data.

point-cloud-filter

Given point cloud data, we apply techniques to separate our object of interest. You can learn more about PCL here.

This is the first perception exercise from Udacity's RoboND.

The scripts showcase the following techniques:

• Downsampling using the Voxel Grid Filter
• Getting the region of interest using a passthrough filter
• Segmentation of the table from everything else using Ransac Plane Fitting
• Reducing noise using statistical outlier filter

Original Point Cloud

Resulting Point Cloud

Dependencies

• Python 2.7, this does not work on Python 3
• PCL bindings by Straw Lab
• PCL tools $ sudo apt-get install pcl-tools
• I used Ubuntu 16.04.2 with ROS full-desktop-version

Viewing Point Cloud Results

• The resulting point clouds can be found in the /point_cloud folder, you can view them with the ff command:

$ pcl_viewer filename.pcd 

Running scripts

• There are three python scripts you can run all of which are commented well.

$ python filename.py

Original Table Scene

Table Scene with Reduced Noise