nicknair/vision-based-navigation

Vision-based-Navigation

Finding the tf of the camera

1. Calibrate the camera

Add the images (10-14) with the chessboard to the calibration folder. Note the K and D matrices and update it in the camera_pose.py script. Make sure to specify resolution of camera in the camera_calibration.py

python3 camera_calibration.py calibration/

2. Getting the camera coordinate with respect to ARUCO marker

Keep the camera in the desired location and run the camera_pose.py script. Note down the tvec and rvec values. This is the camera pose with respect to the marker.

python3 camera_pose.py 

3. Finding the Lidar to Marker transform

Physcially measure the translation and rotation of the marker with respect to the lidar.

4. Adding the tfs in the launch file

Installing and running spatio-temporal voxel layer

1. Installing the package from binaries or souce

2. Adding the config parameters to costmap.yaml

3. Adding the plugin in local_costmap.yaml and/or global_costmap.yaml

4. Adding/Checking the map to camera_aligned_to_depth_color_frame

5. Use NavStack as usual.