读取文件夹内所有视频，然后每秒截一帧图像 python opencv

本文链接：https://blog.youkuaiyun.com/baiyang10010/article/details/103412605

本文介绍了一种使用Python和OpenCV将视频文件分解为单个图像帧的方法。通过遍历指定目录下的所有视频文件，该脚本可以按每24帧抽取一帧的方式将视频转换为图像，有效地进行视频数据的预处理。

摘要生成于 C知道，由 DeepSeek-R1 满血版支持，前往体验 >

#coding=utf-8
import cv2
import numpy as np
#获得文件夹内所有视频的名字
import  os

'''
for each vedio in vedios
    分解视频成图像
'''
SourceImgPath = "C:\\PycharmProjects\\VediosToImgs\\Vedio" + '\\'  # 视频读取路径
vedionamelist = os.listdir(SourceImgPath)  # 获得所有视频名字列表

ImgWritePath = "C:\\PycharmProjects\\VediosToImgs\\Img" + '\\'  # 图像保存路径
img_end = ".jpg"
img_start = 0


for vedio_path in vedionamelist:
    VedioPath = os.path.join('%s%s' % (SourceImgPath, vedio_path))  # 获得文件夹下所有文件的路径   读取路径和保存路径
    cap = cv2.VideoCapture(VedioPath)
    while cap.read():
        # get a frame
        ret, frame = cap.read()
        if ret == False:
            break #读到文件末尾

        # 显示第几帧
        frames_num = cap.get(cv2.CAP_PROP_POS_FRAMES)
        # 显示实时帧数
        FPS = cap.get(cv2.CAP_PROP_FPS)
        # 总帧数
        total_frame = cap.get(cv2.CAP_PROP_FRAME_COUNT)
        # show information of frame
        # cv2.putText(frame, "FPS:"+str(FPS), (17, 13), cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0, 255, 255))
        # cv2.putText(frame, "NUM OF FRAME:"+str(frames_num), (222, 15), cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0, 255, 255))
        # cv2.putText(frame, "TOTAL FRAME:" + str(total_frame), (504, 17), cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0, 255, 255))
        # show a frame
        cv2.imshow("capture", frame)
        # img name
        img_name = str(img_start) + img_end
        img_start = img_start + 1
        # 存储
        if frames_num % 24 == 0:
            cv2.imwrite(ImgWritePath + img_name, frame)

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

    cap.release()
    cv2.destroyAllWindows()