ROS2之OpenCV基础代码对比foxy~galactic~humble

Posted 2022-06-13 zhangrelay

参考：

automaticaddison.com/getting-started-with-opencv-in-ros-2-galactic-python/

automaticaddison.com/getting-started-with-opencv-in-ros-2-foxy-fitzroy-python/

---

推荐使用：YOLOX + ROS2 object detection package

也可以参考：github.com/jeffreyttc/opencv_ros2

vision_opencv
ros2 vision_opencv 包含将 ROS 2 与 OpenCV 接口的包，OpenCV 是一个专为计算效率和实时计算机视觉应用程序而设计的库。 该存储库包含：

1. cv_bridge：ROS 2 图像消息和 OpenCV 图像表示之间的桥梁
2. image_geometry：处理图像和像素几何的方法集合
3. opencv_tests：集成测试以使用带有 opencv 的包的功能
4. vision_opencv：安装 cv_bridge 和 image_geometry 的元包

为了将 ROS 2 与 OpenCV 一起使用，请参阅 cv_bridge 包中的详细信息。

---

程序适用于foxy/galactic/humble，windows/linux系统通用 

---

在本教程中，将学习如何将 ROS 2 与流行的计算机视觉库 OpenCV 接口的基础知识。 这些基础知识将提供在机器人应用程序中添加视觉的基础。

我们将创建一个图像发布者节点以将网络摄像头数据（即视频帧）发布到一个主题，我们将创建一个订阅该主题的图像订阅者节点。

---

pub发布：

foxy

galactic

# Basic ROS 2 program to publish real-time streaming # video from your built-in webcam # Author: # - Addison Sears-Collins    # Import the necessary libraries import rclpy # Python Client Library for ROS 2 from rclpy.node import Node # Handles the creation of nodes from sensor_msgs.msg import Image # Image is the message type from cv_bridge import CvBridge # Package to convert between ROS and OpenCV Images import cv2 # OpenCV library class ImagePublisher(Node):   """   Create an ImagePublisher class, which is a subclass of the Node class.   """   def __init__(self):     """     Class constructor to set up the node     """     # Initiate the Node class's constructor and give it a name     super().__init__('image_publisher')            # Create the publisher. This publisher will publish an Image     # to the video_frames topic. The queue size is 10 messages.     self.publisher_ = self.create_publisher(Image, 'video_frames', 10)            # We will publish a message every 0.1 seconds     timer_period = 0.1  # seconds            # Create the timer     self.timer = self.create_timer(timer_period, self.timer_callback)               # Create a VideoCapture object     # The argument '0' gets the default webcam.     self.cap = cv2.VideoCapture(0)               # Used to convert between ROS and OpenCV images     self.br = CvBridge()       def timer_callback(self):     """     Callback function.     This function gets called every 0.1 seconds.     """     # Capture frame-by-frame     # This method returns True/False as well     # as the video frame.     ret, frame = self.cap.read()                if ret == True:       # Publish the image.       # The 'cv2_to_imgmsg' method converts an OpenCV       # image to a ROS 2 image message       self.publisher_.publish(self.br.cv2_to_imgmsg(frame))     # Display the message on the console     self.get_logger().info('Publishing video frame')    def main(args=None):      # Initialize the rclpy library   rclpy.init(args=args)      # Create the node   image_publisher = ImagePublisher()      # Spin the node so the callback function is called.   rclpy.spin(image_publisher)      # Destroy the node explicitly   # (optional - otherwise it will be done automatically   # when the garbage collector destroys the node object)   image_publisher.destroy_node()      # Shutdown the ROS client library for Python   rclpy.shutdown()    if __name__ == '__main__':   main()

# Basic ROS 2 program to publish real-time streaming # video from your built-in webcam # Author: # - Addison Sears-Collins     # Import the necessary libraries import rclpy # Python Client Library for ROS 2 from rclpy.node import Node # Handles the creation of nodes from sensor_msgs.msg import Image # Image is the message type from cv_bridge import CvBridge # Package to convert between ROS and OpenCV Images import cv2 # OpenCV library    class ImagePublisher(Node):   """   Create an ImagePublisher class, which is a subclass of the Node class.   """   def __init__(self):     """     Class constructor to set up the node     """     # Initiate the Node class's constructor and give it a name     super().__init__('image_publisher')             # Create the publisher. This publisher will publish an Image     # to the video_frames topic. The queue size is 10 messages.     self.publisher_ = self.create_publisher(Image, 'video_frames', 10)             # We will publish a message every 0.1 seconds     timer_period = 0.1  # seconds             # Create the timer     self.timer = self.create_timer(timer_period, self.timer_callback)                # Create a VideoCapture object     # The argument '0' gets the default webcam.     self.cap = cv2.VideoCapture(0)                # Used to convert between ROS and OpenCV images     self.br = CvBridge()        def timer_callback(self):     """     Callback function.     This function gets called every 0.1 seconds.     """     # Capture frame-by-frame     # This method returns True/False as well     # as the video frame.     ret, frame = self.cap.read()                 if ret == True:       # Publish the image.       # The 'cv2_to_imgmsg' method converts an OpenCV       # image to a ROS 2 image message       self.publisher_.publish(self.br.cv2_to_imgmsg(frame))        # Display the message on the console     self.get_logger().info('Publishing video frame')     def main(args=None):       # Initialize the rclpy library   rclpy.init(args=args)       # Create the node   image_publisher = ImagePublisher()       # Spin the node so the callback function is called.   rclpy.spin(image_publisher)       # Destroy the node explicitly   # (optional - otherwise it will be done automatically   # when the garbage collector destroys the node object)   image_publisher.destroy_node()       # Shutdown the ROS client library for Python   rclpy.shutdown()     if __name__ == '__main__':   main()

---

sub订阅：

foxy

galactic

# Basic ROS 2 program to subscribe to real-time streaming # video from your built-in webcam # Author: # - Addison Sears-Collins    # Import the necessary libraries import rclpy # Python library for ROS 2 from rclpy.node import Node # Handles the creation of nodes from sensor_msgs.msg import Image # Image is the message type from cv_bridge import CvBridge # Package to convert between ROS and OpenCV Images import cv2 # OpenCV library class ImageSubscriber(Node):   """   Create an ImageSubscriber class, which is a subclass of the Node class.   """   def __init__(self):     """     Class constructor to set up the node     """     # Initiate the Node class's constructor and give it a name     super().__init__('image_subscriber')            # Create the subscriber. This subscriber will receive an Image     # from the video_frames topic. The queue size is 10 messages.     self.subscription = self.create_subscription(       Image,       'video_frames',       self.listener_callback,       10)     self.subscription # prevent unused variable warning            # Used to convert between ROS and OpenCV images     self.br = CvBridge()       def listener_callback(self, data):     """     Callback function.     """     # Display the message on the console     self.get_logger().info('Receiving video frame')     # Convert ROS Image message to OpenCV image     current_frame = self.br.imgmsg_to_cv2(data)          # Display image     cv2.imshow("camera", current_frame)          cv2.waitKey(1)    def main(args=None):      # Initialize the rclpy library   rclpy.init(args=args)      # Create the node   image_subscriber = ImageSubscriber()      # Spin the node so the callback function is called.   rclpy.spin(image_subscriber)      # Destroy the node explicitly   # (optional - otherwise it will be done automatically   # when the garbage collector destroys the node object)   image_subscriber.destroy_node()      # Shutdown the ROS client library for Python   rclpy.shutdown()    if __name__ == '__main__':   main()

# Basic ROS 2 program to subscribe to real-time streaming # video from your built-in webcam # Author: # - Addison Sears-Collins     # Import the necessary libraries import rclpy # Python library for ROS 2 from rclpy.node import Node # Handles the creation of nodes from sensor_msgs.msg import Image # Image is the message type from cv_bridge import CvBridge # Package to convert between ROS and OpenCV Images import cv2 # OpenCV library    class ImageSubscriber(Node):   """   Create an ImageSubscriber class, which is a subclass of the Node class.   """   def __init__(self):     """     Class constructor to set up the node     """     # Initiate the Node class's constructor and give it a name     super().__init__('image_subscriber')             # Create the subscriber. This subscriber will receive an Image     # from the video_frames topic. The queue size is 10 messages.     self.subscription = self.create_subscription(       Image,       'video_frames',       self.listener_callback,       10)     self.subscription # prevent unused variable warning             # Used to convert between ROS and OpenCV images     self.br = CvBridge()        def listener_callback(self, data):     """     Callback function.     """     # Display the message on the console     self.get_logger().info('Receiving video frame')        # Convert ROS Image message to OpenCV image     current_frame = self.br.imgmsg_to_cv2(data)           # Display image     cv2.imshow("camera", current_frame)           cv2.waitKey(1)     def main(args=None):       # Initialize the rclpy library   rclpy.init(args=args)       # Create the node   image_subscriber = ImageSubscriber()       # Spin the node so the callback function is called.   rclpy.spin(image_subscriber)       # Destroy the node explicitly   # (optional - otherwise it will be done automatically   # when the garbage collector destroys the node object)   image_subscriber.destroy_node()       # Shutdown the ROS client library for Python   rclpy.shutdown()     if __name__ == '__main__':   main()

---

humble：

pub

# Basic ROS 2 program to publish real-time streaming

# video from your built-in webcam

# Author:

# - Addison Sears-Collins

   

# Import the necessary libraries

import rclpy # Python Client Library for ROS 2

from rclpy.node import Node # Handles the creation of nodes

from sensor_msgs.msg import Image # Image is the message type

from cv_bridge import CvBridge # Package to convert between ROS and OpenCV Images

import cv2 # OpenCV library

  

class ImagePublisher(Node):

  """

  Create an ImagePublisher class, which is a subclass of the Node class.

  """

  def __init__(self):

    """

    Class constructor to set up the node

    """

    # Initiate the Node class's constructor and give it a name

    super().__init__('image_publisher')

       

    # Create the publisher. This publisher will publish an Image

    # to the video_frames topic. The queue size is 10 messages.

    self.publisher_ = self.create_publisher(Image, 'video_frames', 10)

       

    # We will publish a message every 0.1 seconds

    timer_period = 0.1  # seconds

       

    # Create the timer

    self.timer = self.create_timer(timer_period, self.timer_callback)

          

    # Create a VideoCapture object

    # The argument '0' gets the default webcam.

    self.cap = cv2.VideoCapture(0)

          

    # Used to convert between ROS and OpenCV images

    self.br = CvBridge()

    

  def timer_callback(self):

    """

    Callback function.

    This function gets called every 0.1 seconds.

    """

    # Capture frame-by-frame

    # This method returns True/False as well

    # as the video frame.

    ret, frame = self.cap.read()

           

    if ret == True:

      # Publish the image.

      # The 'cv2_to_imgmsg' method converts an OpenCV

      # image to a ROS 2 image message

      self.publisher_.publish(self.br.cv2_to_imgmsg(frame))

  

    # Display the message on the console

    self.get_logger().info('Publishing video frame')

   

def main(args=None):

   

  # Initialize the rclpy library

  rclpy.init(args=args)

   

  # Create the node

  image_publisher = ImagePublisher()

   

  # Spin the node so the callback function is called.

  rclpy.spin(image_publisher)

   

  # Destroy the node explicitly

  # (optional - otherwise it will be done automatically

  # when the garbage collector destroys the node object)

  image_publisher.destroy_node()

   

  # Shutdown the ROS client library for Python

  rclpy.shutdown()

   

if __name__ == '__main__':

  main()

 sub

# Basic ROS 2 program to subscribe to real-time streaming

# video from your built-in webcam

# Author:

# - Addison Sears-Collins

   

# Import the necessary libraries

import rclpy # Python library for ROS 2

from rclpy.node import Node # Handles the creation of nodes

from sensor_msgs.msg import Image # Image is the message type

from cv_bridge import CvBridge # Package to convert between ROS and OpenCV Images

import cv2 # OpenCV library

  

class ImageSubscriber(Node):

  """

  Create an ImageSubscriber class, which is a subclass of the Node class.

  """

  def __init__(self):

    """

    Class constructor to set up the node

    """

    # Initiate the Node class's constructor and give it a name

    super().__init__('image_subscriber')

       

    # Create the subscriber. This subscriber will receive an Image

    # from the video_frames topic. The queue size is 10 messages.

    self.subscription = self.create_subscription(

      Image,

      'video_frames',

      self.listener_callback,

      10)

    self.subscription # prevent unused variable warning

       

    # Used to convert between ROS and OpenCV images

    self.br = CvBridge()

    

  def listener_callback(self, data):

    """

    Callback function.

    """

    # Display the message on the console

    self.get_logger().info('Receiving video frame')

  

    # Convert ROS Image message to OpenCV image

    current_frame = self.br.imgmsg_to_cv2(data)

     

    # Display image

    cv2.imshow("camera", current_frame)

     

    cv2.waitKey(1)

   

def main(args=None):

   

  # Initialize the rclpy library

  rclpy.init(args=args)

   

  # Create the node

  image_subscriber = ImageSubscriber()

   

  # Spin the node so the callback function is called.

  rclpy.spin(image_subscriber)

   

  # Destroy the node explicitly

  # (optional - otherwise it will be done automatically

  # when the garbage collector destroys the node object)

  image_subscriber.destroy_node()

   

  # Shutdown the ROS client library for Python

  rclpy.shutdown()

   

if __name__ == '__main__':

  main()

---