Calibrated Camera rig json file properties.

Autonomous Vehicles DRIVE PX 2 DriveWorks
1

I have performed camera calibration according to the real guidelines. Below is the results of the generated rig.json file. Where do I obtain the camera’s extrinsic properties such as roll, pitch and yaw on the vehicle. There is two parameters which are “correction_sensor_R_FLU” and “nominalSensor2Rig_FLU”. Which one do i use?

{
    "rig": {
        "sensors": [
            {
                "correction_rig_T": [
                    0.0,
                    0.0,
                    0.0
                ],
                "correction_sensor_R_FLU": {
                    "roll-pitch-yaw": [
                        1.77590813166262e-08,
                        9.09948312255437e-08,
                        6.15486384081265e-10
                    ]
                },
                "name": "camera-0",
                "nominalSensor2Rig_FLU": {
                    "roll-pitch-yaw": [
                        -0.488398611545563,
                        1.69403576850891,
                        -0.00453641591593623
                    ],
                    "t": [
                        1.96111714839935,
                        -0.0420609898865223,
                        1.55862772464752
                    ]
                },
                "parameter": "",
                "properties": {
                    "Model": "ocam",
                    "c": "1.0053",
                    "cx": "960.00000",
                    "cy": "604.00000",
                    "d": "0.00000",
                    "e": "0.00000",
                    "height": "1208",
                    "poly": "1.88932226562500e+3 -8.69694128632545e-2 -8.61710359458812e-5 -1.62127818725821e-7 -2.04988560847941e-11 ",
                    "width": "1920"
                },
                "protocol": "camera"
            },
        ],
        "vehicle": {
            "valid": false,
            "value": {
                "axlebaseFront": 0.0,
                "axlebaseRear": 0.0,
                "bumperFront": 0.0,
                "bumperRear": 0.0,
                "centerOfMassToRearAxle": 0.0,
                "frontCorneringStiffness": 0.0,
                "height": 0.0,
                "inertia": 0.0,
                "length": 0.0,
                "mass": 0.0,
                "rearCorneringStiffness": 0.0,
                "steeringCoefficient": 0.0,
                "wheelDiameter": 0.0,
                "wheelbase": 0.0,
                "width": 0.0,
                "widthWithMirrors": 0.0
            }
        }
    },
    "version": 2
2

Hi,

1) nominalSensor2Rig_FLU: the nominal sensor to rig transformation.
2) correction_rig_T: shows the corrections in the translation with respect to the nominalSensor2Rig transformation.
3) correction_sensor_R_FLU: shows the corrections in the rotation with respect to the nominalSensor2Rig transformation.

Thanks.

3

Hi, another questions on this topic are that:

  1. what does the word ‘rig’ mean in the calibration?
  2. As it well known to us all that, in the calibration, we will get the intrinsic and extrinsic parameters, but why we need the nominalSensor2Rig_FLU, correction_rig_T, correction_sensor_R_FLU?
  3. In the rig.json, if we want to get the intrinsics of camera0, which parameters?