Most unmanned vehicle systems use depth data to interpret their environments. One popular format is a point cloud, which is an mx3 array of [x, y, z] coordinates corresponding to the locations of measured depth points in 3D space. MathWorks tools can be used to design algorithms which filter and interpret large point cloud data sets. The resulting interpretation can be used in connection with MATLAB ROS+Gazebo or physical modeling tools to simulate a vehicle as it accomplishes a task.
- View the module from the Getting Started Guide "Visualize and Cluster Point Cloud Data"
- Computer Vision System: Use 3D vision functionality to visualize, manipulate, and identify point clouds
- Statistics: Provides statistical and machine learning algorithms that can be used to cluster and identify patterns within point clouds
- Instrument Control Toolbox: Use Serial, UDP, or TCP/IP to receive point cloud data from LIDAR devices. Not supported for code generation.
- DSP System Toolbox: Use UDP to receive point cloud from LIDAR devices. Supported for code generation.
- Image Acquisition: acquire depth video from a Kinect