Skip to content

Latest commit

 

History

History
85 lines (62 loc) · 4.01 KB

image_demo.md

File metadata and controls

85 lines (62 loc) · 4.01 KB
Raw

Image

In this demo, we will make RViz visualize an image. This means we have to publish a sensor_msgs::msg::Image message to a topic that we set the plugin to subscribe to.

We can find ROS1 the documentation for the message type here, or the ROS2 specification on GitHub.

We could probably figure out how assemble this message our selves, but thankfully there's the cv_bridge library which does all the heavy lifting for us. Nevertheless, we still have to do some coding.

In this demo, I'm assuming that you're reading your images using OpenCV. If you aren't you need to figure out how to convert it from your data type, into a cv::Mat format. After that you're in the clear!

C++

The gist of this demo is:

  1. Load the image using OpenCV
  2. Create a sensor_msgs::msg::Image message and populate it using the cv_bridge library.
  3. Publish the message to a topic, e.g. lena.

The code can be found here.

CMakeLists.txt

In order to compile, you need to make sure you have the following in your CMakeLists.txt.

find_package(rclcpp REQUIRED)
find_package(cv_bridge REQUIRED)
find_package(sensor_msgs REQUIRED)

add_executable(image_demo
  src/image_demo.cpp
)
# Specify dependencies for this specific target
ament_target_dependencies(image_demo
  rclcpp
  cv_bridge
  sensor_msgs
)

# Install in /base_ws/install/ so that `ros2 run` can find it. 
install(TARGETS
  image_demo
  DESTINATION lib/${PROJECT_NAME}
)

Compile

colcon build --packages-select rviz_demo

Running

(Note: Remember to source the local installation: source install/local_setup.bash)

Run RViz:

ros2 run rviz2 rviz2

Then add the image plugin, and specify the topic name as lena. And finally you can run the demo package:

ros2 run rviz_demo image_demo

In RViz, you should now see a progression like this:

(Btw: If you are not familiar with the picture of Lena Forsén I recommend you to read the Wikipedia article about it. It has a long standing history in the Computer Vision community. Similar to how the Utah pot and the Stanford bunny is famous in the Computer Graphics community.

Troubleshooting

If you get strange garbled errors like below, then you have forgotten to add a dependency under ament_target_dependencies in the CMakeLists.txt. In the example below, it's because I forgot to add sensor_msgs.

CMakeFiles/image_demo.dir/src/image_demo.cpp.o: In function `rclcpp::Publisher<sensor_msgs::msg::Image_<std::allocator<void> >, std::allocator<void> >::Publisher(rclcpp::node_interfaces::NodeBaseInterface*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, rclcpp::QoS const&, rclcpp::PublisherOptionsWithAllocator<std::allocator<void> > const&)':
image_demo.cpp:(.text._ZN6rclcpp9PublisherIN11sensor_msgs3msg6Image_ISaIvEEES4_EC2EPNS_15node_interfaces17NodeBaseInterfaceERKNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEERKNS_3QoSERKNS_29PublisherOptionsWithAllocatorIS4_EE[_ZN6rclcpp9PublisherIN11sensor_msgs3msg6Image_ISaIvEEES4_EC5EPNS_15node_interfaces17NodeBaseInterfaceERKNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEERKNS_3QoSERKNS_29PublisherOptionsWithAllocatorIS4_EE]+0x6a): undefined reference to `rosidl_message_type_support_t const* rosidl_typesupport_cpp::get_message_type_support_handle<sensor_msgs::msg::Image_<std::allocator<void> > >()'

Questions

When I compile, I get the following warning:

/usr/bin/ld: warning: libopencv_imgcodecs.so.3.2, needed by /opt/ros/eloquent/lib/libcv_bridge.so, may conflict with libopencv_imgcodecs.so.4.4
/usr/bin/ld: warning: libopencv_core.so.3.2, needed by /opt/ros/eloquent/lib/libcv_bridge.so, may conflict with libopencv_core.so.4.4

I am not sure how to mitigate these.