Provides a C++ script to fuse two point clouds from two stereo cameras.

• PCL
• Eigen

Place point cloud files pointcloud_fusion/data/pcd_pos1/pcd_2.pcd and pointcloud_fusion/data/pcd_pos2/pcd_2.pcd.

To fuse the two point clouds, one needs the transformation estimated by the lidar_camera_calibration find_transformation.launch node, which is a rotation and translation as a 4x4 matrix that transform all points in the LiDAR frame to the camera frame.

In pointcloud_fusion/data/rts.txt provide,
T1, transformation from LiDAR-to-Camera1 (corresponding .pcd file in pcd_pos1) T2, transformation from LiDAR-to-Camera2 (corresponding .pcd file in pcd_pos2)

Go to pointcloud_fusion/build and compile the program.

cmake ..
make

If you encounted an error, something that looks like this,

CMake Error: The current CMakeCache.txt directory /home/user-name/lidar_camera_calibration/pointcloud_fusion/build/CMakeCache.txt is different than the directory /home/user-name-2/lidar_camera_calibration/pointcloud_fusion/build/CMakeCache.txt where CMakeCache.txt was created. This may result in binaries being created in the wrong place. If you are not sure, reedit the CMakeCache.txt

then, delete the CMakeCache.txt and compile the program.

Run

./fusion

to fuse the point clouds.

The fused point cloud will be saved in pointcloud_fusion/data as final_cloud.pcd and can be viewed using pcl_viewer.