Point Cloud Segmentation with PCL and ROS

Euclidean cluster based point cloud segmentation of a cluttered desk environment using PCL and ROS. RGBA data is simulated in a Gazebo environment. ROS segmentation node uses PCL to perform voxel downsampling and passthrough filtration to reduce point cloud size, RANSAC planar model fitting to remove the table, and euclidean cluster extraction to identify individual objects. Final clusters are color coded to denote difference. Execution time is <.2s on standard laptop processor, useful as a perception node in networked ROS environment.

Experience with RGBA sensors, Point Clouds, PCL, ROS, C++, Euclidean Cluster Extraction, RANSAC Model Fitting, OOP. 

Full code and description available here