Point Grey (FLIR) Spinnaker based camera driver (Blackfly S etc.)
- Maintainer status: maintained
- Maintainer: Vikrant Shah <vikrantshah+ros AT gmail DOT com>
- Author: Abhishek Bajpayee, Pushyami Kaveti, Vikrant Shah
- License: MIT
- Source: git https://github.com/neufieldrobotics/spinnaker_sdk_camera_driver.git (branch: master)
These are the ros drivers for running the Pt Grey (FLIR) cameras that use the Spinnaker SDK. This code has been tested with various Point Grey Blackfly S (BFS) cameras.
These instructions will get you a copy of the project up and running on your local machine for development and testing purposes.
The pre-requisites for this repo include:
- ros-kinetic-image-transport * libunwind-dev
# after installing spinnaker verify that you can run your cameras with SpinView # after installing ros, install other pre-requisites with: sudo apt install libunwind-dev ros-kinetic-cv-bridge ros-kinetic-image-transport
To install the spinnaker drivers:
bash mkdir -p ~/spinnaker_ws/src cd spinnaker_ws/src git clone https://github.com/neufieldrobotics/spinnaker_sdk_camera_driver.git cd ~/spinnaker_ws/ catkin_make source ~/spinnaker_ws/devel/setup.bash # add this to ~/.bashrc to make this permanent
Running the drivers
Modify the params/test_params.yaml file replacing the cam-ids and master cam serial number to match your camera's serial number. Then run the code as:
roslaunch spinnaker_camera_driver acquisition.launch # Test that the images are being published by running rqt_image_view
acquisition_nodeThe node that runs the spinakker sdk and exports to ROS. All the parameters can be set via the launch file or via the yaml config_file. It is good practice to specify all the 'task' specific parameters via launch file and all the 'system configuration' specific parameters via a config_file.
Published Topicscamera_array/cam0/image_raw (sensor_msgs/Image)
- Depending on number of cameras connected you would have that many publishers for image messages
- Depending on number of cameras connected you would have that many publishers for camera info messages if you provide the camera calibration information
Parameters~binning (int, default: 1)
- Binning for cameras, when changing from 2 to 1 cameras need to be unplugged and replugged
- Should color images be used (only works on models that support color images)
- Exposure setting for cameras, 0 is auto-exposure
- Number of frames to save/view 0=ON
- Show images on screen GUI
- Show images on screen GUI in a grid
- Flag whether images should be saved or not (via opencv mat objects to disk)
- Location to save the image data
- Type of file type to save to when saving images locally: binary, tiff, bmp, jpeg etc.
- When hybrid software triggering is used, this controls the FPS, 0=as fast as possible
- Show time/FPS on output
- Flag whether images should be published to ROS. When manually selecting frames to send to rosbag, set this to False. In that case, frames will only be sent when 'space bar' is pressed
- Flag whether each image should have Unique timestamps vs the master cams time stamp for all
- Flag for max rate mode which is when the master triggers the slaves and saves images at maximum rate possible. This is the multithreaded mode"
- This is a list of camera serial numbers in the order which it would be organized. The convention is to start from left to right.
- This is the names that would be given to the cameras for filenames and rostopics in the order specified above eg. cam0, cam1 etc.
- This is the serial number of the camera to be used as master for triggering the other cameras
- Number of frames to be skipped initially to flush the buffer
- Seconds to wait in the deinit/init sequence
- Distortion model for the camera calibration.
- Distortion coefficients of all the cameras in the array. Must match the number of cam_ids provided.
- Intrinsic coefficients of all the cameras in the array. Must match the number of cam_ids provided. Specified as [fx 0 cx 0 fy cy 0 0 1]