EECE_VSLAM

This uses StellaSLAM to build 3d maps of an environment for a drone. Do this stuff: https://chat.openai.com/share/d1563399-f824-4829-9f36-a67f7969c753

Getting Started

Do the makefile pip set up stuff. Install ROS Foxy.

You should be able to just run

colcon build --packages-select drone_stella_slam source foxy_ws/install/setup.bash

./scripts/InstallStella.sh

In theory the below should just work...

This seems to work, but you need to use the parameter bridge described in the drive and use the scripts/bridge.yaml file instead of all 1to2 topics for it not to be laggy.

Make sure you recursively clone the submodules for stella vslam ros: git submodule update --init --recursive

To visualize the image data you need to change the type from map to the frame one under it

Each top level bullet is a terminal

1. cd scripts
   1. ./setup.sh
      1. This will install all dependencies
2. source noetic
   1. douglas@douglasvm:~/foxy_ws$ roslaunch realsense2_camera rs_camera.launch
3. Source noetic THEN source foxy
   1. rosparam load bridge.yaml
   2. ros2 run ros1_bridge parameter_bridge
4. source foxy
   1. ros2 run image_transport republish raw in:=camera/depth/image_rect_raw out:=/camera/depth/image_raw
5. source foxy
   1. cd src/drone_stella_slam
   2. export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/home/foxy_vm/MOTION-STELLA-VSLAM/stella_build/stella_vslam/build/lib/
   3. ros2 run stella_vslam_ros run_slam -v orb_vocab.fbow -c camera_config/realsense_rgbd.yaml --map-db-out map.msg

OLD STUFF

Each top level bullet is a terminal

1. source noetic
   1. Run roscore
2. ~/foxy_ws$ rosparam load scripts/bridge.yaml
   1. douglas@douglasvm:~/foxy_ws$ roslaunch realsense2_camera rs_camera.launch
3. Source noetic THEN source foxy
   1. rosparam load scripts/bridge.yaml
   2. ros2 run ros1_bridge parameter_bridge
      1. jk that doesn't seem to work just send it all: ros2 run ros1_bridge --bridge-all-1to2-topics
4. source foxy
   1. ros2 run image_transport republish raw in:=camera/depth/image_rect_raw out:=/camera/depth/image_raw
5. source foxy_ws (this assumes you've built it already)
   1. cd src
   2. ros2 run stella_vslam_ros run_slam -v orb_vocab.fbow -c camera_config/realsense_rgbd.yaml --map-db-out map.msg

Open up 3 terminals Terminal 1: ros2 launch realsense2_camera rs_launch.py Terminal 2: ros2 run image_transport republish raw in:=camera/color/image_raw out:=/camera/image_raw Terminal 2 (depth realsense): ros2 run image_transport republish raw in:=camera/depth/image_rect_raw out:=/camera/depth/image_raw Mapping:

Terminal 3: ros2 run stella_vslam_ros run_slam -v orb_vocab.fbow -c camera_config/equirectangular.yaml --map-db-out map.msg Terminal 3 (realsense): ros2 run stella_vslam_ros run_slam -v orb_vocab.fbow -c camera_config/realsense_rgbd.yaml --map-db-out map.msg Localization:

Terminal 3: ros2 run stella_vslam_ros run_slam --disable-mapping -v orb_vocab.fbow -c camera_config/equirectangular.yaml --map-db-in map.msg Terminal 3 (realsense): ros2 run stella_vslam_ros run_slam --disable-mapping -v orb_vocab.fbow -c camera_config/realsense_rgbd.yaml --map-db-in map.msg