Opencv VideoWriter:如何获得与模型的实时推理具有相同“速率”的输出视频?

以下是一个代码,可以在mp4视频中“实时”了解YOLO模型的推断:

import cv2
from ultralytics import YOLO

model = YOLO('yolov8n.pt')

video_path = "path/to/your/video/file.mp4"
cap = cv2.VideoCapture(video_path)

while cap.isOpened():
    success, frame = cap.read()
    if success:
        results = model(frame)
        annotated_frame = results[0].plot()
        cv2.imshow("YOLOv8 Inference", annotated_frame)
        if cv2.waitKey(1) & 0xFF == ord("q"):
            break
    else:
        break
cap.release()
cv2.destroyAllWindows()

当我运行上面的代码时,它会打开VLC并显示带有框等的视频。由于模型需要时间来进行预测,所以它相当滞后。

我想要这个结果(滞后的视频),但是 已记录 。

问题是,如果我这样做:

import cv2
from ultralytics import YOLO

model = YOLO('yolov8n.pt')

video_path = "path/to/your/video/file.mp4"
cap = cv2.VideoCapture(video_path)

frame_num = int(cap.get(cv2.CAP_PROP_FRAME_COUNT))
frame_width = int(cap.get(cv2.CAP_PROP_FRAME_WIDTH))
frame_height = int(cap.get(cv2.CAP_PROP_FRAME_HEIGHT))
fourcc = cv2.VideoWriter_fourcc(*'mp4v')
video_writer = cv2.VideoWriter('video_opencv.mp4',fourcc,30,(frame_width,frame_height))

while cap.isOpened():
    success, frame = cap.read()
    if success:
        results = model(frame)
        annotated_frame = results[0].plot()
        #cv2.imshow("YOLOv8 Inference", annotated_frame)
        video_writer.write(annotated_frame)
        if cv2.waitKey(1) & 0xFF == ord("q"):
            break
    else:
        break
cap.release()
cv2.destroyAllWindows()

这基本上是一样的,但我录制视频而不是简单地显示它,然后当我阅读视频时,它的速度和原始视频一样快(没有预测)

我希望我是清白的。

有什么解决办法吗?