Datasets:
Camera Intrinsics for AirZoo Training Data
#3
by zouzhengyu77 - opened
Hi, thank you for releasing AirZoo.
I found RGB, depth, and camera poses in the released data, but could not find the camera intrinsics. Could you please provide (f_x, f_y, c_x, c_y), or clarify whether the same intrinsics are used for all sequences?
Thank you!
Hi,
All sequences share the same pinhole K at a given resolution (no distortion):
1600×1200: (1931.7, 1931.7, 800, 600), FOV 45.0° × 34.5°
1920×1080: (2317.6, 2317.6, 960, 540), FOV 45.0° × 26.2°
Use the pair matching the RGB size.
Best regards