Skip to content
This repository has been archived by the owner on Jan 24, 2018. It is now read-only.

Latest commit

 

History

History
181 lines (119 loc) · 5.24 KB

README.md

File metadata and controls

181 lines (119 loc) · 5.24 KB
Raw

ROS Node for Intel® RealSense™ SLAM Library

This package contains a ROS wrapper for Intel's SLAM library. The realsense_ros_slam package provides a solution for SLAM as a ROS nodelet. It consumes the messages sent by the realsense_ros_camera nodelet, and publishes messages for the camera pose and occupancy map.

Hardware/Software Requirements

To use realsense_ros_slam, you need a mobile agent with a RealSense ZR300 camera.

Ubuntu requirements:

  • Ubuntu 16.04
  • gcc 4.9.3
  • libeigen3-dev
  • libc++ dev

Required Intel libraries:

  • Linux SDK
  • librealsense_slam

Inputs and Outputs

Subscribed Topics

camera/fisheye/image_raw

  • Message type: sensor_msgs::Image
  • The fisheye image with timestamp

camera/depth/image_raw

  • Message type: sensor_msgs::Image
  • The depth image with timestamp

camera/fisheye/camera_info

  • Message type: sensor_msgs::CameraInfo
  • The intrinsics of the fisheye camera

camera/depth/camera_info

  • Message type: sensor_msgs::CameraInfo
  • The intrinsics of the depth camera

camera/gyro/sample

  • Message type: sensor_msgs::Imu
  • Gyroscope sample with timestamp

camera/accel/sample

  • Message type: sensor_msgs::Imu
  • Accelerometer sample with timestamp

camera/gyro/imu_info

  • Message type: realsense_ros_camera::IMUInfo
  • Gyroscope intrinsics, noise and bias variances

camera/accel/imu_info

  • Message type: realsense_ros_camera::IMUInfo
  • Accelerometer intrinsics, noise and bias variances

camera/extrinsics/fisheye2imu

  • Message type: realsense_ros_camera::Extrinsics
  • Fisheye to IMU extrinsics

camera/extrinsics/fisheye2depth

  • Message type: realsense_ros_camera::Extrinsics
  • Fisheye to depth extrinsics

Published Topics

camera_pose

  • Message type: geometry_msgs::PoseStamped
  • The raw camera pose, in the camera's coordinate system (right-handed, +x right, +y down, +z forward)

reloc_pose

  • Message type: geometry_msgs::PoseStamped
  • The relocalized camera pose, in the camera's coordinate system. Published only when a relocalization has occurred.

pose2d

  • Message type: geometry_msgs::Pose2D
  • The 2D camera pose, projected onto a plane corresponding to the occupancy map

tracking_accuracy

  • Message type: realsense_ros_slam::TrackingAccuracy
  • The current 6DoF tracking accuracy (low/medium/high/failed). Currently high is not used.

map

  • Message type: nav_msgs::OccupancyGrid
  • The occupancy map

Parameters

trajectoryFilename

  • str::string, default: 'trajectory.ppm'
  • The name of the trajectory file to output. The files will be saved in the realsense_ros_slam directory.

relocalizationFilename

  • str::string, default: 'relocalization.bin'
  • The name of re-localization file to output. The files will be saved in the realsense_ros_slam directory.

occupancyFilename

  • str::string, default: 'occupancy.bin'
  • The name of occupancy data file to output.

map_resolution

  • double
  • Sets the size of the grid squares in the occupancy map, in meters.

hoi_min

  • double
  • Sets the minimum height of interest for the occupancy map, in meters.

hoi_max

  • double
  • Sets the maximum height of interest for the occupancy map, in meters.

doi_min

  • double
  • Sets the minimum depth of interest for the occupancy map, in meters.

doi_max

  • double
  • Sets the maximum depth of interest for the occupancy map, in meters.

Services

realsense_ros_slam/reset

  • Resets SLAM so that all subsequent poses are relative to the current position.
  • Clears the occupancy map.

Usage

To run the slam engine:

$ cd catkin_ws
$ source devel/setup.bash
$ roslaunch realsense_ros_slam demo_slam.launch

To run the slam engine using a recorded bag file:

$ roslaunch realsense_ros_slam demo_slam_from_bag.launch bag_path:=~/test.bag

To record slam topics to a bag file:

$ roslaunch realsense_ros_slam record_bag_slam.launch

The bag file will be written to your home directory, with a timestamp appended.

To print the estimated pose messages, in another console window:

$ rostopic echo pose2d

The demo_slam.launch and demo_slam_from_bag.launch files will automatically start rviz using a configuration file located at launch/demo_settings.rviz. The raw camera pose, occupancy map, and odometry are shown in rviz. The odometry message is only sent by the SLAM nodelet for demo purposes, since the pose2d message cannot be displayed by rviz. The odometry message contains the same 2D pose that the pose2d message does. This shows where the robot is located relative to the occupancy map.

Testing

The SLAM package can be tested with pre-recorded data using the provided ROS unit test. No physical camera needs to be present in order to run the test. The following steps can be used to build the unit test and download the pre-recorded ROS .bag data:

$ cd ~/catkin_ws
$ catkin_make -DREALSENSE_ENABLE_TESTING=On
$ rostest realsense_ros_slam slam.test

You will see the test execute on the command line as it builds a map from the input data, then the test passes with a "RESULT: SUCCESS" status.

Note: Depending on your internet connection speed, enabling 'REALSENSE_ENABLE_TESTING' can cause catkin_make to run for very long time (more than 5 minutes), as it downloads required pre-recorded .bag data.