ROS Package For LeadSense Stereo Camera

This package lets you use the LeadSense stereo camera with ROS. It outputs the rectified left and right images, depth map, disparity, point cloud, IMU information and obstacle detection result.

Prerequisites

• Ubuntu 16.04
• EvoBinoSDK >= 1.7.0 and its dependency CUDA
• ROS Kinetic

Note: if you want to use LeadSense without GPU, you could edit leadsense_ros_nodelet.cpp: change header file from evo_depthcamera.h to evo_stereocamera.h, change the camera class from DepthCamera to StereoCamera, and remove evo::MAT_TYPE_CPU in retrieve functions (CPU version do not have this option).

Note: if you want to visualize disparity, you will also need to install the image_view package:

sudo apt-get install ros-kinetic-image-view

Note: if you are using a LeadSense with IMU and want to visualize the IMU information using RVIZ, you will also need to install the IMU plugin:

sudo apt-get install ros-kinetic-rviz-imu-plugin

Note: if you want to dynamic configure the parameters with GUI, you will also need to install the rqt_reconfigure package:

sudo apt-get install ros-kinetic-rqt-reconfigure

Quick Start

1. Create a workspace named catkin_ws under your HOME folder following the ros tutorial

2. Open a terminal, download and build the package:

cd ~/catkin_ws/src
git clone https://github.com/ShanghaiEyevolutionTechnology/leadsense_ros.git
cd ../
catkin_make
echo source $(pwd)/devel/setup.bash >> ~/.bashrc
source ~/.bashrc

3. Start the LeadSense camera:

with RVIZ:

roslaunch leadsense_ros leadsense_rviz.launch 

without RVIZ:

roslaunch leadsense_ros leadsense_nodisplay.launch 

Note: you could set the parameters by adding options of the launch. Example:

• If you just want to get images and do not need depth calculation, add calc_distance:=false.
• If you want to do obstacle detection, add obstacle_detection:=true.

Topics

The topics provided will be introduced.

Rectified Image Topics

• /leadsense/left/image_rect_gray
• /leadsense/right/image_rect_gray

To visualize a image, you can use the simple Image plugin since the image data are published on topics of type sensor_msgs/Image.

Depth Topic

• /leadsense/depth/depth_registered

To visualize a depth image, you can use the simple Image plugin since the depth data are published on topics of type sensor_msgs/Image. The data of depth image is encoded in float (32-bit). Unclick Normalize Range, then set distance range with Min value and Max value for a good visualization result (unit: m).

Point Cloud Topic

• /leadsense/point_cloud/cloud_registered

To directly visualize a topic of type sensor_msgs/Pointcloud2 you can use the Pointcloud2 plugin.

Disparity Topic

• /leadsense/disparity/disparity

Use the disparity_view node in the image_view package to visualize it:

rosrun image_view disparity_view image:=/leadsense/disparity/disparity

IMU Topics

Fusion result and raw data topics are provided. When you choose 9 axes mode, fusion is done by 9 axes, and magnet topic is published. When you choose 6 axes mode, fusion is done by 6 axes.

• /leadsense/imu/data

Fusion result and calibrated data. To directly visualize a topic of type sensor_msgs/Imu, you can use the IMU plugin. The orientation and axes will be shown.

• /leadsense/imu/data_raw

Raw data do not have orientation information, so you can not visualize it. You can check the data by:

rostopic echo /leadsense/imu/data_raw

• /leadsense/imu/mag
• /leadsense/imu/mag_raw

Magnet data is published on a topic of type sensor_msgs/MagneticField. You can check the data by:

rostopic echo /leadsense/imu/mag
rostopic echo /leadsense/imu/mag_raw

Obstacle Detection Topics

Note: there must be ground in the image, or obstacle detection can not be done.

When obstacle detection is enable, these topics are provided:

• /leadsense/left/camera_height
• /leadsense/left/camera_pitch
• /leadsense/left/camera_roll

Camera position information are published on topics of type std_msgs/Float32. You can check the data by:

rostopic echo /leadsense/left/camera_height
rostopic echo /leadsense/left/camera_pitch
rostopic echo /leadsense/left/camera_roll

• /leadsense/left/laser_scan

To visualize a laser scan, you can use the simple LaserScan plugin since the image data are published on topics of type sensor_msgs/LaserScan.

• /leadsense/left/image_with_obstacle
• /leadsense/left/image_with_ground

To visualize a image, you can use the simple Image plugin since the image data are published on topics of type sensor_msgs/Image.

Dynamic reconfigure

You can dynamically change many configuration parameters during the execution:

rosrun rqt_reconfigure rqt_reconfigure