如何提取边界框的 x 和 y 坐标

Question

我的代码定义了一个 ROS 节点，用于使用 YOLO 算法在图像中进行对象检测，特别是从指定文件路径加载的 YOLO 模型。该节点订阅了广播图像的 ROS 主题 (/camera/color/image_raw)。收到图像后，会触发回调函数，使用 CvBridge 将 ROS 图像消息转换为 OpenCV 图像格式。然后该图像由 YOLO 模型处理以检测对象。结果（包括检测到的对象框）将打印到控制台。为了防止连续处理和重新处理图像，在处理第一个图像后设置一个标志（image_processed）。此外，处理后的图像将保存到指定的文件夹（临时）。该节点保持活动状态，侦听新图像，直到手动关闭。

#!/usr/bin/python3
import rospy
from sensor_msgs.msg import Image
from cv_bridge import CvBridge
import cv2
import os
from ultralytics import YOLO

class Node(object):
    def __init__(self):
        self.yolo_model = YOLO('/home/user/catkin_ws/src/run_folder/content/runs/obb/train/weights/best.pt')
        self.br = CvBridge()
        self.image_processed = False  # Flag to indicate if the image has been processed

        # Subscribe to the ROS topic that publishes images
        rospy.Subscriber("/camera/color/image_raw", Image, self.callback)

    def callback(self, msg):
        if not self.image_processed:  # Process only if no image has been processed yet
            image = self.br.imgmsg_to_cv2(msg)  # Convert ROS Image message to OpenCV image

            # Perform object detection with YOLO
            results = self.yolo_model.predict(image, show=True)
            for r in results[0]:
                print("---------------------------")
                print(r.boxes)

            # Save the image after processing it for the first time
            self.save_image(image, 'temporary')
            self.image_processed = True  # Set the flag to True to avoid re-processing

    def save_image(self, image, folder_name):
        if not os.path.exists(folder_name):
            os.makedirs(folder_name)
        
        file_path = os.path.join(folder_name, 'image.jpg')
        cv2.imwrite(file_path, image)
        print(f"Image saved at {file_path}")

if __name__ == '__main__':
    rospy.init_node("image_processor_with_yolo", anonymous=True)
    node = Node()
    rospy.spin()  # Keep the node running until shutdown

当我尝试提取边界框时，我在输出中得到“无”。

Answer 1

与Yolo8：

for r in results:
    box = result.boxes.xyxy
    x1 = round(box[0][0].item())
    y1 = round(box[0][1].item())
    x2 = round(box[0][2].item())
    y2 = round(box[0][3].item())