Tensorflow 2.0 视频分类(二) UCF-101数据集预处理

关于UCF-101介绍请参看：
https://blog.csdn.net/Forrest97/article/details/105938520

UCF-101数据下载

视频下载路径：http://crcv.ucf.edu/data/UCF101/UCF101.rar
解压后就是分类数据集的标准目录格式，二级目录名为人类活动类别，二级目录下就是对应的视频数据。每个视频长度为4s，大小320*240， 帧率25HZ。
需要注意: 相同的活动下，有不同的视频是截取自同一个长视频的片段，即视频中的人物和背景等特征基本相似，因此为避免此类视频被分别划分到train和test集合引起训练效果不合实际的过大，UCF放提供了标准的train和test集合检索文件。
集合划分检索文件下载地址：
https://www.crcv.ucf.edu/data/UCF101/UCF101TrainTestSplits-RecognitionTask.zip
解压后如下，有三组推荐的划分方案

建立Train 和 Test 数据集合目录

运行代码1_move_files.py
将解压后的UCF-101和ucfTrainTestlist放在相同目录下运行以下代码，完成对原有视频的重新划分和移动

import os
import os.path

def get_train_test_lists(version='01'):
    """
    Using one of the train/test files (01, 02, or 03), get the filename
    breakdowns we'll later use to move everything.
    选择一个数据分割版本，并读取检索路径
    """
    # Get our files based on version. 
    test_file = os.path.join('ucfTrainTestlist', 'testlist' + version + '.txt')
    train_file = os.path.join('ucfTrainTestlist', 'trainlist' + version + '.txt')

    # Build the test list.
    with open(test_file) as fin:
        test_list = [row.strip() for row in list(fin)]

    # Build the train list. Extra step to remove the class index.
    with open(train_file) as fin:
        train_list = [row.strip() for row in list(fin)]
        train_list = [row.split(' ')[0] for row in train_list]

    # Set the groups in a dictionary.
    file_groups = {
        'train': train_list,
        'test': test_list
    }

    return file_groups

def move_files(file_groups):
    """This assumes all of our files are currently in _this_ directory.
    So move them to the appropriate spot. Only needs to happen once.
    将视频文件移动到新建的路径中
    """
    # Do each of our groups.
    for group, videos in file_groups.items():

        # Do each of our videos.
        for video in videos:

            # Get the parts.
            #parts = video.split(os.path.sep)
            parts = video.split('/') #修改
            classname = parts[0]
            filename = parts[1]

            # Check if this class exists.
            if not os.path.exists(os.path.join(group, classname)):
                print("Creating folder for %s/%s" % (group, classname))
                os.makedirs(os.path.join(group, classname))

            # Check if we have already moved this file, or at least that it
            # exists to move
            filename_input=os.path.join('UCF-101',classname, filename) #新增加
            if not os.path.exists(filename_input):
                print("Can't find %s to move. Skipping." % (filename))
                continue

            # Move it.
            dest = os.path.join(group, classname, filename)
            print("Moving %s to %s" % (filename, dest))
            os.rename(filename_input, dest)

    print("Done.")

def main():
    """
    Go through each of our train/test text files and move the videos
    to the right place.
    """
    os.makedirs('test')
    os.makedirs('train')
    # Get the videos in groups so we can move them.
    group_lists = get_train_test_lists()

    # Move the files.
    move_files(group_lists)

if __name__ == '__main__':
    main()

ffmpeg提取视频中的画面

ffmpeg具体的安装可参考：https://blog.csdn.net/www588555/article/details/105135644
其中注意环境变量增加后，如果命令行能调用，但是在Python中出现如下错误：
‘ffmpeg’ 不是内部或外部命令，也不是可运行的程序
可考虑重启以下电脑或者
call([“ffmpeg”, “-i”, src, dest])替换为call([“path of ffmpeg.exe”, “-i”, src, dest])

"""
After moving all the files using the 1_ file, we run this one to extract
the images from the videos and also create a data file we can use
for training and testing later.
"""
import csv
import glob
import os
import os.path
from subprocess import call

def extract_files():
    """After we have all of our videos split between train and test, and
    all nested within folders representing their classes, we need to
    make a data file that we can reference when training our RNN(s).
    This will let us keep track of image sequences and other parts
    of the training process.

    We'll first need to extract images from each of the videos. We'll
    need to record the following data in the file:

    [train|test], class, filename, nb frames

    Extracting can be done with ffmpeg:
    `ffmpeg -i video.mpg image-%04d.jpg`
    """
    data_file = []
    folders = ['train', 'test']

    for folder in folders:
        class_folders = glob.glob(os.path.join(folder, '*'))

        for vid_class in class_folders:
            class_files = glob.glob(os.path.join(vid_class, '*.avi'))

            for video_path in class_files:
                # Get the parts of the file.
                video_parts = get_video_parts(video_path)

                train_or_test, classname, filename_no_ext, filename = video_parts

                # Only extract if we haven't done it yet. Otherwise, just get
                # the info.
                if not check_already_extracted(video_parts):
                    # Now extract it.
                    src = os.path.join(train_or_test, classname, filename)
                    dest = os.path.join(train_or_test, classname,
                        filename_no_ext + '-%04d.jpg')
                    call(["ffmpeg", "-i", src, dest])

                # Now get how many frames it is.
                nb_frames = get_nb_frames_for_video(video_parts)

                data_file.append([train_or_test, classname, filename_no_ext, nb_frames])

                print("Generated %d frames for %s" % (nb_frames, filename_no_ext))
	#生成csv文件保存视频信息
    with open('data_file.csv', 'w') as fout:
        writer = csv.writer(fout)
        writer.writerows(data_file)

    print("Extracted and wrote %d video files." % (len(data_file)))

def get_nb_frames_for_video(video_parts):
    """Given video parts of an (assumed) already extracted video, return
    the number of frames that were extracted."""
    train_or_test, classname, filename_no_ext, _ = video_parts
    generated_files = glob.glob(os.path.join(train_or_test, classname,
                                filename_no_ext + '*.jpg'))
    return len(generated_files)

def get_video_parts(video_path):
    """Given a full path to a video, return its parts."""
    parts = video_path.split(os.path.sep)
    filename = parts[2]
    filename_no_ext = filename.split('.')[0]
    classname = parts[1]
    train_or_test = parts[0]

    return train_or_test, classname, filename_no_ext, filename

def check_already_extracted(video_parts):
    """Check to see if we created the -0001 frame of this file."""
    #查看是否已经生成图片
    train_or_test, classname, filename_no_ext, _ = video_parts
    return bool(os.path.exists(os.path.join(train_or_test, classname,
                               filename_no_ext + '-0001.jpg')))

def main():
    """
    Extract images from videos and build a new file that we
    can use as our data input file. It can have format:

    [train|test], class, filename, nb frames
    """
    os.makedirs('sequences')
    extract_files()

if __name__ == '__main__':
    main()

这个过程时间较长，耐心等待。

参考

https://blog.coast.ai/five-video-classification-methods-implemented-in-keras-and-tensorflow-99cad29cc0b5
https://github.com/harvitronix/five-video-classification-methods