yolov5检测视频流

是的，YOLOv5 可以用于检测视频流。您可以使用 Python 的 OpenCV 库来捕获视频流，然后将每一帧图像输入到 YOLOv5 模型中进行目标检测。以下是一个简单的示例代码：

import cv2
import torch
from models.experimental import attempt_load
from utils.general import non_max_suppression, scale_coords, plot_one_box
from utils.torch_utils import select_device

# Load YOLOv5 model
weights = 'yolov5s.pt'
device = select_device('')

model = attempt_load(weights, map_location=device)
model.to(device).eval()

# Initialize video capture
cap = cv2.VideoCapture(0)

# Define detection threshold
conf_threshold = 0.5

while True:
    # Read frame from video capture
    ret, frame = cap.read()

    # Convert BGR image to RGB
    frame = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)

    # Resize image to YOLOv5 input size
    img = cv2.resize(frame, (640, 640))

    # Normalize image
    img = img / 255.0

    # Convert image to PyTorch tensor
    img = torch.from_numpy(img).float().to(device)
    img = img.permute(2, 0, 1).unsqueeze(0)

    # Run YOLOv5 inference
    detections = model(img)[0]

    # Apply NMS
    detections = non_max_suppression(detections, conf_threshold)

    # Plot bounding boxes and labels on original frame
    for detection in detections:
        if detection is not None:
            detection = scale_coords(img.shape[2:], detection[:, :4], frame.shape).round()
            for x1, y1, x2, y2, conf, cls in detection:
                label = f'{model.names[int(cls)]} {conf:.2f}'
                plot_one_box((x1, y1, x2, y2), frame, label=label)

    # Display frame
    cv2.imshow('frame', cv2.cvtColor(frame, cv2.COLOR_RGB2BGR))

    # Press q to exit
    if cv2.waitKey(1) & 0xFF == ord('q'):
        break

# Release video capture and destroy windows
cap.release()
cv2.destroyAllWindows()

请注意，此代码仅适用于使用 OpenCV 捕获的实时视频流。如果您要检测现有的视频文件，则需要在代码中进行一些修改。