Any Method about getting extrinsic matrix of each depth frame?

[wangmiaowei]

In my understanding about Azure Kinect, is there any function here to get extrinsic matrix of each depth frame? Maybe the first frame is set as the global pose and then use IMU data to compute the pose of other frames? I do not known whether it is possbile?

[dasparli]

You can get the extrinsic matrices for AKDK through this function which is part of the API: Azure Kinect Sensor SDK: k4a_device_t::k4a_device_get_calibration (microsoft.github.io)

You are able to use these matrices for mapping/transformation between the depth/color/IMU (there are functions built in for this: Use Azure Kinect calibration functions | Microsoft Docs). There is no “extrinsic matrix for each frame” unless you are pointing the Azure Kinect at some sort of target (such as in this example: Azure-Kinect-Sensor-SDK/examples/calibration_registration at develop · microsoft/Azure-Kinect-Sensor-SDK (github.com)) and recalculating the extrinsics on each frame.