This is the vision package of the Hamburg Bit-Bots. For standardized cameras such as USB-webcams we use the wolves_image_provider package as image provider. Alternatively every image source, that publishes sensor_msgs/Image messages (i.e. basler driver for basler cameras) is supported. Settings considering the vision are set in the visionparams.yaml (bitbots_vision/config/visionparams.yaml). You do NOT want to enable DEBUG on the robot. The color calibration files are created with the wolves colorpicker and a rosbag. The source code of the vision is located in bitbots_vision/src. In bitbots_vision/models the fcnn models are stored. Due to their size, these are not part of this repository. After changes in the models or config/color_spaces directory, rebuilding the vision node is required to see these changes in dynamic reconfigure. To tweak the camera image, use the settings in the image provider.
pkg="bitbots_vision" type="vision.py" name="bitbots_vision"
pub_config topic:'/speak' type:Speak speak_publisher topic:'vision_config' type:Config pub_balls topic: config['ROS_ball_msg_topic'] type:BallsInImage pub_lines topic:config['ROS_line_msg_topic'] type:LineInformationInImage pub_obstacle topic:config['ROS_obstacle_msg_topic'] type:ObstaclesInImage pub_goal topic:config['ROS_goal_msg_topic'] type:GoalInImage pub_ball_fcnn topic:config['ROS_fcnn_img_msg_topic'] type:ImageWithRegionOfInterest pub_debug_image topic:config['ROS_debug_image_msg_topic'] type:Image pub_debug_fcnn_image topic:config['ROS_debug_fcnn_image_msg_topic'] type:Image
image_sub topic: config['ROS_img_msg_topic'] type:Image callback: _image_callback head_sub topic: config['ROS_head_joint_msg_topic'] type:JointState callback: _head_joint_state_callback
pkg="bitbots_vision" type="dynamic_color_space.py" name="bitbots_dynamic_color_space"
pub_color_space topic: vision_config['ROS_dynamic_color_space_msg_topic'] type:ColorSpace
sub_vision_config_msg topic: 'vision_config' type:Config callback: vision_config_callback sub_image_msg topic: vision_config['ROS_img_msg_topic']type:Image callback: image_callback
Currently, the models of our neural networks are not available publicly.
Due to their size, they are not included in the repository.
Bit-Bots find them in the Mafiasi NextCloud robocup-ai/log/
In the bitbots_vision_tools directory, special tools for debugging/introspection purposes are provided.
To start the vision, use
roslaunch bitbots_vision vision_startup.launch
sim:=true does activate simulation time, switch to simulation color settings and deactivate launching of an image provider
camera:=false does deactivate all image providers (e.g. for use with rosbags or in simulation)
basler:=falsedoes start wolves image provider instead of the basler camera driver
dummyball:=true does not start the ball detection to save resources
debug:=true does activate publishing of several debug images which can be inspected in the rqt image view
use_game_settings:=true does load additional game settings
bitbots_vision
- vision_startup: starts the vision and a camera image provider
- Params:
- sim [true/FALSE],
- camera [TRUE/false],
- basler [TRUE/false],
- dummyball [true/FALSE],
- debug [true/FALSE],
- use_game_settings [true/FALSE]
- Params:
bitbots_vision_tools
- bitbots_vision_recordbag: records a rosbag containing the camera image as image_raw and stores it as /tmp/tmpbag.bag
clone bitbots_vision and usb_cam in your bitbots_mate
- compiling
Delete the deve, build, logs, and.catkin_tools folders in your workspace，and run
catkin config -DPYTHON_EXECUTABLE=/usr/bin/python3
catkin build
You need to perform a few operations to run the simulation environment,after you source the path,of course
roslaunch bitbots_vision vision_startup.launch sim:=true camera:=false