Ironsides - ROS Integration

This package lets you use the Ironsides visual inertial computing module with ROS.

Getting started

• Install ROS (Kinetic)
  • http://wiki.ros.org/kinetic/Installation/Ubuntu
• Create a ROS Workspace
  • mkdir -p ~/catkin_ws/src
  • cd ~/catkin_ws/
  • catkin_make
  • source ~/catkin_ws/devel/setup.bash
• Import and make our code
  • place pirvs_ros folder in ~/catkin_ws/src
  • run init python script to import PIRVS SDK
    • cd ~/catkin_ws/src/pirvs_ros/init
    • python pirvs_ros_init.py [PIRVS SDK folder] [calib_raw JSON] [calib_rectified JSON]
  • cd ~/catkin_ws
  • catkin_make

Quick Launch

• Try Stereo_camera_node
  • switch Ironsides to Raw Mode
  • roslaunch pirvs_ros raw_camera.launch
• Try Depth_node
  • switch Ironsides to Depth Mode
  • roslaunch pirvs_ros depth_camera.launch
• Try Slam_node
  • switch Ironsides to Raw Mode
  • roslaunch pirvs_ros display.launch

Test Stereo_camera_node

• Published topic:
  • StereoImage/left: left image
  • StereoImage/right: right image
  • IMU/data_raw: IMU
  • StereoImage/left/cameraInfo: left cam calibration
  • StereoImage/right/cameraInfo: right cam calibration
• Run Stereo_camera_node:
  • rosrun pirvs_ros stereo_camera_node [cam_calib_left.yaml] [cam_calib_right.yaml]
• Check image:
  • rosrun image_view image_view image:=[topic name]
  • e.g., rosrun image_view image_view image:=/pirvs_cam_raw/StereoImage/left
• Check IMU,cameraInfo:
  • rostopic echo [topic name]
  • e.g., rostopic echo IMU/data_raw

Test Depth_camera_node

• Published topic:
  • StereoImage/rect_left: rectified left image
  • StereoImage/rect_right: rectified right image
  • StereoImage/detph: depth image
• Run Depth_camera_node:
  • rosrun pirvs_ros depth_camera_node
• Check image:
  • rosrun image_view image_view image:=[topic name]
  • e.g., rosrun image_view image_view image:=/pirvs_cam_depth/StereoImage/rect_left

Test Slam_node

• Published topic:
  • published topics:
  • StereoImage/left: left image
  • StereoImage/right: right image
  • Pose: camera pose
  • odom: odom information for kviz
• Run slam_node:
  • rosrun pirvs_ros slam_node [calib.json] [vocab.json]
• Check image:
  • rosrun image_view image_view image:=[topic name]
  • e.g., rosrun image_view image_view image:=/pirvs_slam/StereoImage/left
• Check Pose:
  • rostopic echo [topic name]
  • e.g., rostopic echo Pose
• Check Pose using RVIZ:
  • rosrun rviz rviz
  • add odometry, listen topic "odom"
  • set fixed frame to my frame

For more information visit PerceptIn ROS Wiki