I have an ASUS Xtion Pro Live that I want to calibrate and use to guide a robot to perform a process that involves touching surfaces. I need to calibrate this camera and use its intrinsic parameters (e.g focal lengths).
I have noticed that running the calibration process and producing a calibration file in .ros/camera_info will change the published depth camera_info to match the IR image, but the actual depth data itself is not changed. This means that later ROS nodes use a different focal length to project the depth data than is used to create the depth from what I presume is a disparity map from the device.
My question can be posed several ways:
How do I find the camera intrinsics used to compute the depth data?
How do I adjust the openni2 nodes so that I can use the same calibrated focal length for both depth calculation and projection into a point cloud?
Howdy,
I have an ASUS Xtion Pro Live that I want to calibrate and use to guide a robot to perform a process that involves touching surfaces. I need to calibrate this camera and use its intrinsic parameters (e.g focal lengths).
I have noticed that running the calibration process and producing a calibration file in
.ros/camera_info
will change the published depth camera_info to match the IR image, but the actual depth data itself is not changed. This means that later ROS nodes use a different focal length to project the depth data than is used to create the depth from what I presume is a disparity map from the device.My question can be posed several ways:
Thanks