[Darknet ROS] darknet_ros yolov3 코드 이것 저것 수정

github.com/leggedrobotics/darknet_ros

leggedrobotics/darknet_ros

 

1. 여러 머신 또는 한 머신에서 multi object detection 수행하기 

-> .launch 파일 수정해야함 

-> namespace 추가하기

<?xml version="1.0" encoding="utf-8"?>

<launch>
  <!-- Console launch prefix -->
  <arg name="launch_prefix" default=""/>
  <arg name="image" default="/image_raw" />

  <!-- Config and weights folder. -->
  <arg name="yolo_weights_path"          default="$(find darknet_ros)/yolo_network_config/weights"/>
  <arg name="yolo_config_path"           default="$(find darknet_ros)/yolo_network_config/cfg"/>

  <!-- Load parameters -->
  <rosparam command="load" ns="darknet_ros_d/darknet_ros" file="$(find darknet_ros)/config/ros_drone.yaml"/>
  <rosparam command="load" ns="darknet_ros_d/darknet_ros" file="$(find darknet_ros)/config/yolov3-tiny.yaml"/>

  <!-- Start darknet and ros wrapper -->
  <node ns="darknet_ros_d" pkg="darknet_ros" type="darknet_ros" name="darknet_ros" output="screen" launch-prefix="$(arg launch_prefix)">$
    <param name="weights_path"          value="$(arg yolo_weights_path)" />
    <param name="config_path"           value="$(arg yolo_config_path)" />
    <remap from="/image_raw"  to="$(arg image)" />
  </node>

 <!--<node name="republish" type="republish" pkg="image_transport" output="screen"      args="compressed in:=/front_camera/image_raw raw out:=/camera/im$
</launch>

 

2. darkent_ros object detection 실행 후 detection_image pubilsh 하려면 

-> /darknet_ros/config/ros.yaml 파일 수정 

-> 맨 아래 enable_opencv 부분 false로  

image_view:

  enable_opencv: false
  wait_key_delay: 1
  enable_console_output: true

 

3. publish되는 topic 이름들 수정하고 싶을 때

-> /darknet_ros/config/ros.yaml 파일 수정

-> actions, publishers topic 이름 수정

actions:

  camera_reading:
    name: /darknet_ros/check_for_objects_d

publishers:

  object_detector:
    topic: /darknet_ros/found_object_d
    queue_size: 1
    latch: false

  bounding_boxes:
    topic: /darknet_ros/bounding_boxes_d
    queue_size: 1
    latch: false

  detection_image:
    topic: /darknet_ros/detection_image_d/compressed
    queue_size: 1
    latch: true

4. publish되는 detection_image를 compressed_image로 publish하고 싶을 때

-> /include/darknet_ros/YoloObjectDetector.hpp, /src/YoloObjectDetector.cpp,  /config/ros.yaml 3가지 파일 수정

-> YoloObjectDetector.hpp 에 CompressedImage Msg include. 

// ROS
#include <actionlib/server/simple_action_server.h>
#include <geometry_msgs/Point.h>
#include <image_transport/image_transport.h>
#include <ros/ros.h>
#include <sensor_msgs/CompressedImage.h>
#include <sensor_msgs/Image.h>
#include <sensor_msgs/image_encodings.h>
#include <std_msgs/Header.h>

-> YoloObjectDetector.cpp 에 detectionImagePublisher 부분들 수정

-> 여기서 param[1] 부분은 JPEG 압축률, 낮을 수록 압축을 많이 한다 ( == 화질이 낮아진다 ) ==> 전송속도 빨라진다

 detectionImagePublisher_ =
      nodeHandle_.advertise<sensor_msgs::CompressedImage>(detectionImageTopicName, detectionImageQueueSize, detectionImageLatch);

// ~~~~

bool YoloObjectDetector::publishDetectionImage(const cv::Mat& detectionImage) {
  if (detectionImagePublisher_.getNumSubscribers() < 1) return false;
  std::vector<uchar> buff;
  std::vector<int> param = std::vector<int>(2);
  param[0] = CV_IMWRITE_JPEG_QUALITY;
  param[1] = 10;
  cv::imencode(".jpg",detectionImage, buff, param);
  cv::Mat jpegimage = cv::imdecode(buff, CV_LOAD_IMAGE_COLOR);
  cv_bridge::CvImage cvImage;
  cvImage.header.stamp = ros::Time::now();
  cvImage.header.frame_id = "detection_image";
  cvImage.encoding = sensor_msgs::image_encodings::BGR8;
  //cvImage.image = detectionImage;
  cvImage.image = jpegimage;
  detectionImagePublisher_.publish(*cvImage.toCompressedImageMsg());
  ROS_DEBUG("Detection image has been published.");
  return true;
}

-> ros.yaml 에서 publihs하는 detection_image topic 이름 뒤에 /compressed 붙이기

  detection_image:
    topic: /darknet_ros/detection_image_d/compressed
    queue_size: 1
    latch: true