抖音无水印视频抓取与按帧截取图片

代码实现无水印视频抓取并根据视频按帧取图。

你喜欢的抖音美女，每一帧都是壁纸，上码：

import requests
import re
from urllib import request
import cv2
import argparse
import os
import urllib


def get_parse(url):
    resp = requests.get(url)
    web_url = resp.url
    item_id = re.search('/(\d+)/', web_url).group(1).strip()
    print(item_id)
    api = 'https://www.iesdouyin.com/web/api/v2/aweme/iteminfo/?item_ids={}'
    info_api = api.format(item_id)
    resp = requests.get(info_api)
    content = resp.text
    urls = re.findall('"(http.*?)"', content)
    return urls, item_id


def get_need(urls):
    need_urls = []
    for url in urls:
        if 'large' in url:
            need_urls.append(url)
        elif 'playwm' in url:
            need_urls.append(url)
        else:
            print('移除项：', url)
    return need_urls


def get_down(need_urls, item_id):
    for i, url in enumerate(need_urls):
        if 'playwm' in url:
            file_name = down_video(url, i, item_id)
        # else:
        # 用于下载图片
        #     down_pic(url, i, item_id)
    return file_name


def down_video(url, i, item_id):
    headers = {
        'User-Agent': 'Mozilla/5.0 (Linux; Android 6.0; Nexus 5 Build/MRA58N) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/83.0.4103.116 Mobile Safari/537.36'
    }
    # 替换链接参数，使用手机heasers获取无水印视频
    url = re.sub('playwm', 'play', url)
    file_name = str(item_id) + "_" + str(i) + '.mp4'
    # start_time = time.time()
    requests.packages.urllib3.disable_warnings()
    resp = requests.get(url, headers=headers, stream=True, verify=False)
    print(resp.headers['content-length'])
    with open('cache/' + file_name, 'wb') as f:
        for i in resp.iter_content(50):
            if i:
                f.write(i)
    print('视频下载成功', url)
    return file_name


# 图片下载方法，暂不调用
def down_pic(url, i, item_id):
    file_name = str(item_id) + '_' + str(i) + '.jpeg'
    try:
        request.urlretrieve(url, 'cache/' + file_name)
        print('下载成功', url)
    except:
        print('下载异常', url)


def parse_args(file_name):
    parser = argparse.ArgumentParser(description='Process pic')
    parser.add_argument('--input', help='video to process', dest='input', default=None, type=str)
    parser.add_argument('--output', help='pic to store', dest='output', default=None, type=str)
    # default为间隔多少帧截取一张图片
    parser.add_argument('--skip_frame', dest='skip_frame', help='skip number of video', default=10, type=int)
    # input为输入视频的路径 ，output为输出存放图片的路径
    args = parser.parse_args(['--input', 'cache/' + file_name, '--output', 'cache/'])
    return args


def process_video(i_video, o_video, num):
    cap = cv2.VideoCapture(i_video)
    num_frame = cap.get(cv2.CAP_PROP_FRAME_COUNT)
    expand_name = '.jpg'
    if not cap.isOpened():
        print("Please check the path.")
    cnt = 0
    count = 0
    while 1:
        ret, frame = cap.read()
        cnt += 1
        if cnt % num == 0:
            count += 1
            cv2.imwrite(os.path.join(o_video, str(count) + expand_name), frame)

        if not ret:
            break


if __name__ == '__main__':
    # 视频下载部分
    url = input('输入链接：')
    url = url.strip()
    urls, item_id = get_parse(url)
    need_urls = get_need(urls)
    file_name = get_down(need_urls, item_id)
    # 图片截取部分
    args = parse_args(file_name)
    if not os.path.exists(args.output):
        os.makedirs(args.output)
    print('执行截取:')
    print(args)
    process_video(args.input, args.output, args.skip_frame)