- extrinsic calibration between a monocular camera and a 3D Lidar by using planar point to plane constraint
- test the algorithm on real dataset. This is the public dataset from the following repository of SubMishMar
conda env create --file environment.yml
conda activate lidar_camera_calibration
- Download dataset from HERE and extract into ./data repository
tar -xf ./data/dataset.tar.xz -C ./data/
- Run calibration app after building the required environment
python scripts/calibration_app.py --dataset_info_json ./data/dataset/dataset_info.json \
--image_list_path ./data/dataset/image_list.txt \
--cloud_list_path ./data/dataset/cloud_list.txt \
--data_path ./data/dataset/data/
you can check the result of lidar point projection on images at /tmp/projected_image_*.jpg
Disclaimer: the result does not look so great on this test, but I hope this repository can still be a template for a standard camera-lidar calibration project