find_object_2d

The find_object_2d package

README

find-object

Linux Build Status
Build Status
Build Status
Windows Build Status

Standalone

Find-Object project, visit the home page for more information.

ROS1

Install

Binaries:

sudo apt-get install ros-$ROS_DISTRO-find-object-2d

Source:

  • To include xfeatures2d and/or nonfree modules of OpenCV, to avoid conflicts with cv_bridge, build same OpenCV version that is used by cv_bridge. Install it in /usr/local (default).

cd ~/catkin_ws
git clone https://github.com/introlab/find-object.git src/find_object_2d
catkin_make

Run

roscore
# Launch your preferred usb camera driver
rosrun uvc_camera uvc_camera_node
rosrun find_object_2d find_object_2d image:=image_raw

See find_object_2d for more information.

ROS2

Install

Binaries:

To come...

Source:

cd ~/ros2_ws
git clone https://github.com/introlab/find-object.git src/find_object_2d
colcon build

Run

# Launch your preferred usb camera driver
ros2 launch realsense2_camera rs_launch.py

# Launch find_object_2d node:
ros2 launch find_object_2d find_object_2d.launch.py image:=/camera/color/image_raw

# Draw objects detected on an image:
ros2 run find_object_2d print_objects_detected --ros-args -r image:=/camera/color/image_raw
3D Pose (TF)

A RGB-D camera is required. Example with Realsense D400 camera:

# Launch your preferred usb camera driver
ros2 launch realsense2_camera rs_launch.py align_depth.enable:=true

# Launch find_object_2d node:
ros2 launch find_object_2d find_object_3d.launch.py \
   rgb_topic:=/camera/color/image_raw \
   depth_topic:=/camera/aligned_depth_to_color/image_raw \
   camera_info_topic:=/camera/color/camera_info

# Show 3D pose in camera frame:
ros2 run find_object_2d tf_example

See find_object_2d for more information (same parameters/topics are used between ROS1 and ROS2 versions).