1. Format
  2. Read and write camera
  3. Step of calibrating LightStage system


For each multiple cameras dataset, you should have intrinsic parameter and extrinsic parameter.

Intrisic parameter contains K, dist for each camera.

Extrinsic parameter contains R,T for each camera, where RX + T converts point X in world coordinate to camera coordinate.

An example of camera paramter can be found in our demo dataset

The metric of T must be meter in all of our codes.

Read and write camera

from easymocap.mytools.camera_utils import read_cameras, write_camera

# path/intri.yml, path/extri.yml
path = 'xxx' 
cameras = read_cameras(path)
write_camera(outdir, cameras)

Step of calibrating LightStage system

# detect the chessboard
python3 apps/calibration/ ${data} --out ${data}/output --pattern 9,6 --grid 0.10
# check the annotation
python3 apps/annotation/ ${data} --annot chessboard --mode chessboard
python3 apps/calibration/ ${ground} --intri ${ba}/output/intri.yml
python3 apps/calibration/ ${ground} --mode cube --out ${ground} --show --write
python3 apps/calibration/ ${ba} --init ${ground} --out ${ba}/calib-base

If it’s failed to detect chessboard in some frames, you can plot a box and press e to re-detect it.

Table of contents