ME5413 AMR Assignment

compile this package

catkin_make -DCATKIN_WHITELIST_PACKAGES="me5413"

Assignment 2

Task 1 Cartographer

launch command

roslaunch me5413 task1.launch bag_filename:=${HOME}/下载/2dlidar.bag

Task 2 Fast-LIO

clone fast-lio to src folder and compile

cd ~/catkin_ws/src
git clone https://github.com/hku-mars/FAST_LIO.git
git clone https://github.com/Livox-SDK/livox_ros_driver.git

cd FAST_LIO
git submodule update --init
cd ../..
catkin_make
source devel/setup.bash

get lidar rotation and transformation in imu link

rosrun tf tf_echo imu_link velo_link

launch fast-lio and play rosbag

roslaunch me5413 task2_fastlio.launch
rosbag play --clock YOUR_ROSBAG_FILE

record rosbag

rosbag record /Odometry /tf /tf_static

Task 2 VINS-Fusion

clone vins-fusion to src folder and compile

cd ~/catkin_ws/src
git clone https://github.com/guisoares9/VINS-Fusion

cd ..
catkin_make -DCATKIN_WHITELIST_PACKAGES="vins"

launch rviz and loop fusion node

roslaunch me5413 task2_vinsfusion.launch

rosrun vins node

rosrun vins vins_node /home/biao/catkin_ws/src/ME5413_AMR/config/vins-fusion/kitti_bag.yaml

record rosbag

rosbag record /vins_estimator/odometry /vins_estimator/point_cloud /vins_estimator/camera_pose /tf /tf_static

Task 2 Coordinate transformation code

from_odometry_output_kitti.py:

Change "/aft_mapped_to_init_high_frec" to the topic that needs to be subscribed to odometry.

Change "camera_gray_left" to the coordinate system node which groundtruth is located.

Change "velo_link" to the coordinate system node which the odometry result is located.

Run the py program in the bag and play another bag containing odometry and tf topics. The code will convert the odometry results into the coordinate system where the groundtruth node is located, and output the txt file of kitti format coordinates in the same directory.

from_odometry_output_tum.py:

Change "/aft_mapped_to_init" to the topic that needs to be subscribed to odometry.

Change "camera_gray_left" to the coordinate system node which groundtruth is located.

Change "camera_init" to the coordinate system node which the odometry result is located.

Run the py program in the bag and play another bag containing odometry and tf topics. The code will convert the odometry results into the coordinate system where the groundtruth node is located, and output the txt file of tum format coordinates in the same directory.

Task 2 Eveluation recorded rosbag

run python script read data from rosbag directly

roscd me5413

python scripts/read_tum_from_rosbag.py '/home/biao/catkin_ws/src/ME5413_AMR/output/task2/fast_lio.bag' --topic /Odometry

evo_ape tum have_fun_tum.txt odom_tum.txt -r full --plot --plot_mode xz

or, run the node and play rosbag

rosrun me5413 record_tum.py
rosbag play YOUR_ROSBAG_FILE