Multi-Camera Calibration Data Component

The multi-camera calibration data component stores the rigid transform from a sensor or camera coordinate system to a common reference coordinate system. It is used in multi-sensor setups where the camera calibration data component alone does not fully describe how data should be aligned in a shared world frame.

The component can be attached to image sets, depth maps, and related data. It is typically used together with per-camera camera calibration data components. For RGB-D rigs, an additional RGB-D calibration data component supplies the depth-to-color relation.

Contents

sensorT — rigid transform from the sensor (or camera) coordinate system to the common reference coordinate system (4×4 matrix). Defaults to the identity matrix.

When sensorT is the identity, algorithms rely only on the camera calibration data component. A non-identity sensorT is required for multi-camera rigs and some RGB-D sensors where an extra sensor-to-reference alignment must be applied on top of the per-camera pose.

Typical use

Attach this component to depth or RGB-D image data processed by algorithms such as Depth Map To Point Cloud, Depth Map To Mesh, Texturing, and Range Image Integration. For point-cloud-to-depth-map conversion inside range image integration, the component is created automatically from the point cloud pose.

Each frame still needs its own intrinsics and camera pose via the camera calibration data component; the multi-camera component supplies the additional sensor-to-reference alignment. For RGB-D input, also attach an RGB-D calibration data component when depth and color cameras are not co-located.

Creating or editing

This data component is attached by certain algorithms, no manual way exists currently.