blob_detection

Detects the blobs (regions of similar brightness or color) with cv::SimpleBlobDetector, filtered by their area, circularity, convexity and inertia.

opencv_apps	3.0.0 and later
ROS	ROS One, jazzy, kilted, lyrical and later
OpenCV	any version of the distributions
Group	Features and shapes, the nodes of which publish ~/corners (Point2DArrayStamped), ~/lines (LineArrayStamped), ~/circles (CircleArrayStamped) and ~/contours (ContourArrayStamped, ~/hulls of convex_hull), by the kind of the features

Usage

# ROS 2
ros2 launch opencv_apps blob_detection.launch.xml image:=/camera/image_raw debug_view:=false
# ROS 1
roslaunch opencv_apps blob_detection.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/ballpit.mp4
# terminal 1: publish the video as image in a loop
ros2 run image_publisher image_publisher_node ballpit.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 blob_detection.launch.xml image:=hsv_color_filter/image min_area:=200.0
# terminal 3: run, keep it running
ros2 launch opencv_apps hsv_color_filter.launch.xml h_limit_min:=40 h_limit_max:=70 s_limit_min:=150 s_limit_max:=255 v_limit_min:=120 v_limit_max:=255
# terminal 4: view the result
ros2 run rqt_image_view rqt_image_view /blob_detection/image

The input is the mask of the yellow balls of hsv_color_filter.

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
~/circles	opencv_apps/msg/CircleArrayStamped
~/image	sensor_msgs/msg/Image

Parameters

See also the common parameters.

Name	Type	Default (range)	Changed at run time	Description
blob_color	int	255 (0 - 255)	yes	0 to extract dark blobs and 255 to extract light blobs
filter_by_area	bool	true	yes	Filter the blobs by the area
filter_by_circularity	bool	false	yes	Filter the blobs by the circularity (4*pi*area/perimeter^2)
filter_by_color	bool	true	yes	Filter the blobs by the color of the center
filter_by_convexity	bool	false	yes	Filter the blobs by the convexity (area / area of convex hull)
filter_by_inertia	bool	false	yes	Filter the blobs by the inertia ratio (the minimum / maximum second central moment, the square of minor axis / major axis for an ellipse, 0: line, 1: circle)
max_area	double	100000.0 (0 - 1e+06)	yes	The maximum area (exclusive) of the blobs [pixel]
max_circularity	double	1.0 (0 - 1)	yes	The maximum circularity (exclusive) of the blobs, 1.0 means no limit
max_convexity	double	1.0 (0 - 1)	yes	The maximum convexity (exclusive) of the blobs, 1.0 means no limit
max_inertia_ratio	double	1.0 (0 - 1)	yes	The maximum inertia ratio (exclusive) of the blobs, 1.0 means no limit
max_threshold	double	220.0 (0 - 255)	yes	The maximum threshold (exclusive) to binarize the image
min_area	double	25.0 (0 - 1e+06)	yes	The minimum area (inclusive) of the blobs [pixel]
min_circularity	double	0.8 (0 - 1)	yes	The minimum circularity (inclusive) of the blobs
min_convexity	double	0.95 (0 - 1)	yes	The minimum convexity (inclusive) of the blobs
min_dist_between_blobs	double	10.0 (0 - 1000)	yes	The blobs located closer than min_dist_between_blobs are merged
min_inertia_ratio	double	0.1 (0 - 1)	yes	The minimum inertia ratio (inclusive) of the blobs
min_repeatability	int	2 (1 - 255)	yes	The minimum number of the binarized images in which a blob is detected
min_threshold	double	50.0 (0 - 255)	yes	The minimum threshold to binarize the image
queue_size	int	3	no, restart the node	Specify queue_size of input image subscribers
threshold_step	double	10.0 (1 - 255)	yes	Step of the thresholds from min_threshold to max_threshold

Source

src/nodelet/blob_detection_nodelet.cpp, launch/blob_detection.launch.xml