Python project to generate extrinsics, intrinsics, 2D projections, 3D points, and rendered images. Currently, only checkerboard calibration grid is supported.

How to run

Run project by providing the config file:

python3 generate.py --cfg config/demo.yaml

Additional arguments:

• --output_folder: output folder, default current directory.
• --save_animation: 1 -- save animation, otherwise 0.
• --save_rendering: 1 -- save rendering, otherwise 0.
• --export_to_meshlab: 1 -- save to meshlab file, otherwise 0.

Output:

• In the output_folder/CFG.NAME/ will be saved calibration data (intrinsics, extrinsics, points) in JSON format. NAME is specified in config file.
• If save_animation is 1, then in output_folder/CFG.NAME/animation/ will be saved (1) animation of calibration grid in 3D space with position of cameras, (2) projection of 3D point to each camera.
• If save_rendering is 1, then in export_to_meshlab/CFG.NAME/render/ will be saved rendered images of calibration grid.
• If export_to_meshlab is 1, then in export_to_meshlab/CFG.NAME/meshlab/ will be saved data suitable to import in Meshlab.

Rendered image

Camera positions

2D projections

Generate new config file using implemented function in config/config.py:

python3 -c "from config.config import generateCircularCameras; generateCircularCameras(save_folder='config/')"