PCL Developers blog

Gabe O’Leary

This is my personal page

project:Documentation and tutorials
mentor:Vincent Rabaud (Michael Dixon [Radu B. Rusu, Patrick Bouffard])

About me

I’m a Computer Science and Engineering student at The University of Michigan.

I can be reached by email at oleary.gabe@gmail.com or by snail mail at:
541 Elm St Ann Arbor, MI 48104

My hobbies include:

  • Backpacking/Hiking
  • Coding
  • Debugging
  • Running
  • Writing documentation


  • Go over the list of tutorials, find out what is missing
    • List of completed and un-completed tutorials here.
  • Build a list of interesting classes that require code examples and documentation. For every interesting class:
    • build sample code that shows its usage
    • add usage code snippets to the doxygen documentation
    • write unit tests
    • [optional] verify performance
  • Reorganize samples, tutorials, tests by modules. Put link in doxygen modules to tutorial modules.
  • Move PCD files into a different directory (repository)
  • Reorganize and introduce “how to” simple tutorials

Click here for a detailed roadmap

Recent status updates

Thursday, August 18, 2011

I’m supposed to add an example to the Statistical removal tutorial that use pcl::PointXYZRGB instead of pcl::PointXYZ, but the input file provided contains no RGB data, so I need a different file to work with for the example. I’ve also been requested to provide an example of finding the volume of irregularly-shaped objects from PointClouds (such as leaves on a bush), and am now just waiting on the files to be provided so that I can write the example.

KdTree Search tutorial complete
Thursday, August 04, 2011

So I finished up the KdTree tutorial, and added a theoretical primer to it to explain a little bit about whats going on. This addition should be live on the tutorial page as soon as the website updates.

KdTree search tutorial.
Tuesday, August 02, 2011

At the moment I’m working on a tutorial for the KdTree module. It is nearly done, I just need to finish up the theoretical primer and then it will be completly good to go. I’ve also gotten several emails regarding the tutorials that I have already written which I need to sit down and respond to. I believe one is about the RANSAC tutorial and the other has to do with my concave/convex hull tutorial.

RANSAC tutorial done moving on
Tuesday, July 26, 2011

At last the Random Sample Consensus tutorial is done. It has 2 seperate examples in it, one using the plane sample consensus model, and the other using the spherical sample consensus model. I was trying to use a cylindrical model, but as I have not yet learned how to generate normals to surfaces I was unable to continue, so I just went with the spherical model instead. Now I am moving on, and my mentor, Vincent, suggested that work on a tutorial for either the Range Images module or the KDTree module, so that is what I am going to begin working on.

Random Sample Consensus tutorial
Tuesday, July 19, 2011

So I finally got the PCLVisualizer demo to work. I just ended up completely removing all of the Point Cloud Library components from my computer and re-compiling and installing the trunk. I believe that the problem I had been encountering was because I had installed some pcl-developer tools by apt-get. Anyway I’m now able to run the tutorial without any issues. This is going to be helpful for making my tutorials becuase I’ll be able to make pictures to include in my documents. I have a lot of the code for my Random Sample Consensus tutorial written, but I’m running into a few issues with displaying it so that everything makes sense. I think what I have written works and demonstrates the algorithm, I’m just working on some pictures that show what’s going on.