ported over cv model #37
Conversation
src/cvmodel_package/CMakeLists.txt
Outdated
| @@ -0,0 +1,43 @@ | |||
| cmake_minimum_required(VERSION 3.8) | |||
| project(cvmodel_package) | |||
There was a problem hiding this comment.
can we rename the package to detection
There was a problem hiding this comment.
including changing folder name, etc
src/cvmodel_package/CMakeLists.txt
Outdated
| target_compile_features(video_publisher PUBLIC c_std_99 cxx_std_17) | ||
|
|
||
| # Add cvmodel_node executable | ||
| add_executable(cvmodel_node src/cvmodel_node.cpp) |
There was a problem hiding this comment.
rename to detection_node as well
|
|
||
| Mat edges = applyCanny(frame); | ||
|
|
||
| frame = drawCenters(frame, "blue"); |
There was a problem hiding this comment.
rather than just draw the centers, we should also publish the results to a different topic. implement this using detections2darray http://docs.ros.org/en/noetic/api/vision_msgs/html/msg/Detection2DArray.html, you only need to populate the position of the pose
| last_frame_time = this->now(); | ||
|
|
||
| } catch (cv_bridge::Exception &e) { | ||
| RCLCPP_ERROR(this->get_logger(), "cv_bridge exception: %s", e.what()); |
There was a problem hiding this comment.
what is this supposed to catch for exactly?
| int frame_height = frame.rows; | ||
| double fps = 120.0; | ||
|
|
||
| writer.open(output_video, VideoWriter::fourcc('X', 'V', 'I', 'D'), fps, Size(frame_width, frame_height)); |
There was a problem hiding this comment.
add in a parameter for video writer. in the standard use case, we don't want to output to video
| } | ||
| } | ||
|
|
||
| void check_for_inactivity() { |
There was a problem hiding this comment.
generally this won't be needed in the future and doesn't fit with ROS but I think it's fine for now
Package Name: cvmodel_package
Ported over python cv_model code to detect lighted up sections. We require a input.mp4, the path/name must be specified in video_publisher.cpp. An output.mp4 is generated, an example of which is below.
output.mp4