lk_flow

Tracks the feature points with the sparse Lucas-Kanade optical flow (cv::calcOpticalFlowPyrLK()). The points are given by the services, or by the corner detection.

opencv_apps	1.11 and later (ROS 1), 3.0.0 and later (ROS 2)
ROS	noetic, ROS One, jazzy, kilted, lyrical and later
OpenCV	any version of the distributions
Group	Optical flow and motion, the nodes of which publish `~/flows` (FlowArrayStamped), `~/shift` (Point2DStamped) of phase_corr

Usage

# ROS 2
ros2 launch opencv_apps lk_flow.launch.xml image:=/camera/image_raw debug_view:=false
# ROS 1
roslaunch opencv_apps lk_flow.launch image:=/camera/image_raw debug_view:=false

Try it with the demo video

Download the input video of the demo above, publish it as image with image_publisher, and run the node on it, in the terminals:

wget https://ros-perception.github.io/opencv_apps/media/clips/keybridge.mp4
# terminal 1: publish the video as image in a loop
ros2 run image_publisher image_publisher_node keybridge.mp4 --ros-args -p publish_rate:=10.0 -r image_raw:=image
# terminal 2: run the node with the parameters of the demo
ros2 launch opencv_apps lk_flow.launch.xml image:=image_rect_throttle use_camera_info:=false
# terminal 3: run, keep it running
ros2 run topic_tools throttle messages image 5 image_rect_throttle
# terminal 4: then run once
ros2 service call /lk_flow/initialize_points std_srvs/srv/Empty
# terminal 5: view the result
ros2 run rqt_image_view rqt_image_view /lk_flow/image

The input is throttled to 5 Hz, and the points to track are initialized by the service.

See Credits of the demo videos for the license of the video.

Subscribed topics

Topic	Type
`image`	`sensor_msgs/msg/Image`

Published topics

Topic	Type
`~/flows`	opencv_apps/msg/FlowArrayStamped
`~/image`	`sensor_msgs/msg/Image`

Services

Service	Type
`~/delete_points`	`std_srvs/srv/Empty`
`~/initialize_points`	`std_srvs/srv/Empty`
`~/toggle_night_mode`	`std_srvs/srv/Empty`

Parameters

See also the common parameters.

Name	Type	Default (range)	Changed at run time	Description
`block_size`	int	3 (1 - 100)	yes	Size of an average block for computing a derivative covariation matrix over each pixel neighborhood.
`harris_k`	double	0.04 (0.001 - 10)	yes	Free parameter of the Harris detector
`min_distance`	int	10 (1 - 100)	yes	Minimum possible Euclidean distance between the returned corners.
`quality_level`	double	0.01 (0.001 - 10)	yes	Parameter characterizing the minimal acceptance quality of image corners.
`queue_size`	int	3	no, restart the node	Specigy queue_size of input image subscribers

Source

src/nodelet/lk_flow_nodelet.cpp, launch/lk_flow.launch.xml