关于 Python 的 asyncio 协程的问题。

先说需求： 我需要处理读取多个视频， 每个视频间隔一定时间取帧进行图像识别。下面是提取视频的代码， 一开始我是使用多进程实现的。后面想着读取视频应该是磁盘 IO， 是否使用协程合适？查资料说不是原生异步需要使用线程池维护， 但是写下来，发现也没关协程啥事啊？是不是我写的有问题。正确写法应该怎么样，求各位大佬解答。

import cv2
import asyncio
import os
import numpy as np
import time

def read(cap):
    ret, frame = cap.read()
    return ret, frame

def put_frame(queue, video_path,interval=50):
    '''
    queue: 队列
    video_path: 视频路径
    interval： 间隔多少帧取一次图片
    '''
    num = 0
    img_list = []
    cap = cv2.VideoCapture(video_path)
    start = time.time()
    while cap.isOpened():
        cap.set(cv2.CAP_PROP_POS_FRAMES, num*interval)
        ret, frame = read(cap)
        if ret:
            custom_image = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
            cv2.imshow("{0}".format(video_path), frame)
            cv2.waitKey(24)
            custom_image = cv2.resize(
            custom_image, (416, 416), interpolation=cv2.INTER_NEAREST)
            custom_image = custom_image.transpose(2, 0, 1)
            img_list.append(custom_image)
            num += 1
        else:
            arr = np.concatenate(img_list, axis=0)
            # arr = np.ascontiguousarray(arr.flat, dtype=np.float32) / 255.0
            queue.put(arr)
            print("--------------, {0}提取视频完成, 用时：{1}", video_path, (time.time()-start))
            break


video_dir = './video'
loop = asyncio.get_event_loop()
tasks = []
queue = asyncio.Queue(loop=loop, maxsize=10)
for video in os.listdir(video_dir):
    # await asyncio.sleep(0)
    video_path = os.path.join(video_dir, video)
    task = loop.run_in_executor(None, put_frame, queue, video_path)
    tasks.append(task)

loop.run_until_complete(asyncio.wait(tasks))



# 他还会出现这个警告：
C:/Users/rookie/Desktop/20200512/demo3.py:36: RuntimeWarning: coroutine 'Queue.put' was never awaited
  queue.put(arr)
RuntimeWarning: Enable tracemalloc to get the object allocation traceback