Images and Point Clouds

High-volume sensor data, such as images and point clouds, require efficient handling to avoid performance bottlenecks in Slicer. SlicerROS2 provides native C++ data bridges that perform high-speed conversion between ROS 2 messages and Slicer’s internal data structures.

Unlike basic topics that return simple data types (like strings or matrices), image and point cloud subscribers are designed to bridge data directly into Slicer’s visualization nodes (Volumes and Models).

Sensor Data Bridges
Slicer type	ROS type	SlicerROS2 “name”
vtkImageData	sensor_msgs/msg/Image	Image
vtkPolyData	sensor_msgs/msg/PointCloud2	PolyData

Explicit Bridging

To avoid scene pollution, SlicerROS2 uses an Explicit Bridging model for sensor data. By default, a subscriber will receive data but will not visualize it. To visualize the data, you must link the subscriber node to a “Target Node” in the Slicer scene.

For Image subscribers, the target node should be a vtkMRMLScalarVolumeNode.
For PolyData subscribers, the target node should be a vtkMRMLModelNode.

When a target node is linked, the SlicerROS2 logic automatically updates that node’s contents whenever a new ROS 2 message arrives, using high-performance C++ memory copying.

Images

Image subscribers support multiple encodings including mono8, rgb8, rgba8, mono16, and 32FC1.

# 1. Create the subscriber
sub = rosNode.CreateAndAddSubscriberNode('Image', '/camera/image_raw')

# 2. Create a Volume node to display the image
volumeNode = slicer.mrmlScene.AddNewNodeByClass('vtkMRMLScalarVolumeNode', 'LiveCamera')

# 3. Link them together
sub.SetTargetNodeID(volumeNode.GetID())

// 1. Create the subscriber
auto sub = rosNode->CreateAndAddSubscriberNode("Image", "/camera/image_raw");

// 2. Create a Volume node
auto volumeNode = vtkMRMLScalarVolumeNode::SafeDownCast(
  scene->AddNewNodeByClass("vtkMRMLScalarVolumeNode", "LiveCamera"));

// 3. Link them
sub->SetTargetNodeID(volumeNode->GetID());

Point Clouds

Point cloud subscribers convert sensor_msgs/PointCloud2 messages into vtkPolyData. The bridge automatically generates vertex cells, making the point cloud immediately renderable in Slicer’s 3D views.

# 1. Create the subscriber
sub = rosNode.CreateAndAddSubscriberNode('PolyData', '/lidar_points')

# 2. Create a Model node to display the points
modelNode = slicer.mrmlScene.AddNewNodeByClass('vtkMRMLModelNode', 'LiveLidar')
modelNode.CreateDefaultDisplayNodes()

# Optional: Set point size for better visibility
modelNode.GetDisplayNode().SetPointSize(3.0)

# 3. Link them
sub.SetTargetNodeID(modelNode.GetID())

// 1. Create the subscriber
auto sub = rosNode->CreateAndAddSubscriberNode("PolyData", "/lidar_points");

// 2. Create a Model node
auto modelNode = vtkMRMLModelNode::SafeDownCast(
  scene->AddNewNodeByClass("vtkMRMLModelNode", "LiveLidar"));
modelNode->CreateDefaultDisplayNodes();

// 3. Link them
sub->SetTargetNodeID(modelNode->GetID());