Camera calibration dataset / pinhole camera model
JonasBchrt opened this issue · 3 comments
@shaozu thank you for making these datasets available to the public.
We would like to run code on your dataset that supports only the pinhole camera model, not the Mei model. Do you think you could provide a calibration dataset such that we can run kalibr and use the model that we support? Or maybe you have calibrated a pinhole camera model yourself already or it would be very easy for you to do?
@JonasBchrt below please find the parameters for the pinhole model.
left
%YAML:1.0
#left camera calibration
model_type: PINHOLE
camera_name: cam1
image_width: 752
image_height: 480
distortion_parameters:
k1: -0.295269
k2: 0.085067
p1: 0.000352
p2: -0.000219
projection_parameters:
fx: 468.157508
fy: 467.978898
cx: 382.180503
cy: 214.028724
right
%YAML:1.0
#right camera calibration
model_type: PINHOLE
camera_name: cam0
image_width: 752
image_height: 480
distortion_parameters:
k1: -0.2997493046323177
k2: 0.09538277522623304
p1: -3.949561182735437e-05
p2: -0.0003873561868834772
projection_parameters:
fx: 470.5799808597001
fy: 470.8173735966212
cx: 359.6496137621658
cy: 250.2687221608846
@shaozu could you also provide the extrinsics (R,t) between the left and right cameras?
@shaozu could you also provide the extrinsics (R,t) between the left and right cameras?
I guess you can figure it out by looking at the extrinsics between each camera and the common IMU, here:
GVINS-Dataset/data/visensor_parameters/left_camera.yaml
Lines 19 to 31 in d6abe89
GVINS-Dataset/data/visensor_parameters/right_camera.yaml
Lines 19 to 31 in d6abe89