Tensorflow(一) TFRecord生成與讀取

TFRecord生成


一、爲什麼使用TFRecord?

正常情況下我們訓練文件夾經常會生成 train, test 或者val文件夾,這些文件夾內部往往會存着成千上萬的圖片或文本等文件,這些文件被散列存着,這樣不僅佔用磁盤空間,並且再被一個個讀取的時候會非常慢,繁瑣。佔用大量內存空間(有的大型數據不足以一次性加載)。此時我們TFRecord格式的文件存儲形式會很合理的幫我們存儲數據。TFRecord內部使用了“Protocol Buffer”二進制數據編碼方案,它只佔用一個內存塊,只需要一次性加載一個二進制文件的方式即可,簡單,快速,尤其對大型訓練數據很友好。而且當我們的訓練數據量比較大的時候,可以將數據分成多個TFRecord文件,來提高處理效率。

二、 生成TFRecord簡單實現方式

我們可以分成兩個部分來介紹如何生成TFRecord,分別是TFRecord生成器以及樣本Example模塊。

  1. TFRecord生成器
writer = tf.python_io.TFRecordWriter(record_path)
writer.write(tf_example.SerializeToString())
writer.close()

這裏面writer就是我們TFrecord生成器。接着我們就可以通過writer.write(tf_example.SerializeToString())來生成我們所要的tfrecord文件了。這裏需要注意的是我們TFRecord生成器在寫完文件後需要關閉writer.close()。這裏tf_example.SerializeToString()是將Example中的map壓縮爲二進制文件,更好的節省空間。那麼tf_example是如何生成的呢?那就是下面所要介紹的樣本Example模塊了。

  1. Example模塊
    首先們來看一下Example協議塊是什麼樣子的。

message Example {
  Features features = 1;
};

message Features {
  map<string, Feature> feature = 1;
};

message Feature {
  oneof kind {
    BytesList bytes_list = 1;
    FloatList float_list = 2;
    Int64List int64_list = 3;
  }
};

我們可以看出上面的tf_example可以寫入的數據形式有三種,分別是BytesList, FloatList以及Int64List的類型。那我們如何寫一個tf_example呢?下面有一個簡單的例子。

def int64_feature(value):
    return tf.train.Feature(int64_list=tf.train.Int64List(value=[value]))

def bytes_feature(value):
    return tf.train.Feature(bytes_list=tf.train.BytesList(value=[value]))

tf_example = tf.train.Example(
        features=tf.train.Features(feature={
            'image/encoded': bytes_feature(encoded_jpg),
            'image/format': bytes_feature('jpg'.encode()),
            'image/class/label': int64_feature(label),
            'image/height': int64_feature(height),
            'image/width': int64_feature(width)}))

下面我們來好好從外部往內部分解來解釋一下上面的內容。
(1)tf.train.Example(features = None) 這裏的features是tf.train.Features類型的特徵實例。
(2)tf.train.Features(feature = None) 這裏的feature是以字典的形式存在,*key:要保存數據的名字    value:要保存的數據,但是格式必須符合tf.train.Feature實例要求。

三、 生成TFRecord文件完整代碼實例

首先我們需要提供數據集

圖片文件夾


通過圖片文件夾我們可以知道這裏面總共有七種分類圖片,類別的名稱就是每個文件夾名稱,每個類別文件夾存儲各自的對應類別的很多圖片。下面我們通過一下代碼(generate_annotation_json.pygenerate_tfrecord.py)生成train.record。

 

  1. generate_annotation_json.py
# -*- coding: utf-8 -*-
# @Time    : 2018/11/22 22:12
# @Author  : MaochengHu
# @Email   : [email protected]
# @File    : generate_annotation_json.py
# @Software: PyCharm

import os
import json


def get_annotation_dict(input_folder_path, word2number_dict):
    label_dict = {}
    father_file_list = os.listdir(input_folder_path)
    for father_file in father_file_list:
        full_father_file = os.path.join(input_folder_path, father_file)
        son_file_list = os.listdir(full_father_file)
        for image_name in son_file_list:
            label_dict[os.path.join(full_father_file, image_name)] = word2number_dict[father_file]
    return label_dict


def save_json(label_dict, json_path):
    with open(json_path, 'w') as json_path:
        json.dump(label_dict, json_path)
    print("label json file has been generated successfully!")
  1. generate_tfrecord.py
# -*- coding: utf-8 -*-
# @Time    : 2018/11/23 0:09
# @Author  : MaochengHu
# @Email   : [email protected]
# @File    : generate_tfrecord.py
# @Software: PyCharm

import os
import tensorflow as tf
import io
from PIL import Image
from generate_annotation_json import get_annotation_dict

flags = tf.app.flags
flags.DEFINE_string('images_dir',
                    '/data2/raycloud/jingxiong_datasets/six_classes/images',
                    'Path to image(directory)')
flags.DEFINE_string('annotation_path',
                     '/data1/humaoc_file/classify/data/annotations/annotations.json',
                    'Path to annotation')
flags.DEFINE_string('record_path',
                    '/data1/humaoc_file/classify/data/train_tfrecord/train.record',
                    'Path to TFRecord')
FLAGS = flags.FLAGS


def int64_feature(value):
    return tf.train.Feature(int64_list=tf.train.Int64List(value=[value]))


def bytes_feature(value):
    return tf.train.Feature(bytes_list=tf.train.BytesList(value=[value]))


def process_image_channels(image):
    process_flag = False
    # process the 4 channels .png
    if image.mode == 'RGBA':
        r, g, b, a = image.split()
        image = Image.merge("RGB", (r,g,b))
        process_flag = True
    # process the channel image
    elif image.mode != 'RGB':
        image = image.convert("RGB")
        process_flag = True
    return image, process_flag


def process_image_reshape(image, resize):
    width, height = image.size
    if resize is not None:
        if width > height:
             width = int(width * resize / height)
             height = resize
        else:
            width = resize
            height = int(height * resize / width)
        image = image.resize((width, height), Image.ANTIALIAS)
    return image


def create_tf_example(image_path, label, resize=None):
    with tf.gfile.GFile(image_path, 'rb') as fid:
        encode_jpg = fid.read()
    encode_jpg_io = io.BytesIO(encode_jpg)
    image = Image.open(encode_jpg_io)
    # process png pic with four channels 
    image, process_flag = process_image_channels(image)
    # reshape image
    image = process_image_reshape(image, resize)
    if process_flag == True or resize is not None:
        bytes_io = io.BytesIO()
        image.save(bytes_io, format='JPEG')
        encoded_jpg = bytes_io.getvalue()
    width, height = image.size
    tf_example = tf.train.Example(
        features=tf.train.Features(
            feature={
                'image/encoded': bytes_feature(encode_jpg),
                'image/format': bytes_feature(b'jpg'),
                'image/class/label': int64_feature(label),
                'image/height': int64_feature(height),
                'image/width': int64_feature(width)
            }
        ))
    return tf_example


def generate_tfrecord(annotation_dict, record_path, resize=None):
    num_tf_example = 0
    writer = tf.python_io.TFRecordWriter(record_path)
    for image_path, label in annotation_dict.items():
        if not tf.gfile.GFile(image_path):
            print("{} does not exist".format(image_path))
        tf_example = create_tf_example(image_path, label, resize)
        writer.write(tf_example.SerializeToString())
        num_tf_example += 1
        if num_tf_example % 100 == 0:
            print("Create %d TF_Example" % num_tf_example)
    writer.close()
    print("{} tf_examples has been created successfully, which are saved in {}".format(num_tf_example, record_path))


def main(_):
    word2number_dict = {
        "combinations": 0,
        "details": 1,
        "sizes": 2,
        "tags": 3,
        "models": 4,
        "tileds": 5,
        "hangs": 6
    }
    images_dir = FLAGS.images_dir
    #annotation_path = FLAGS.annotation_path
    record_path = FLAGS.record_path
    annotation_dict = get_annotation_dict(images_dir, word2number_dict)
    generate_tfrecord(annotation_dict, record_path)


if __name__ == '__main__':
    tf.app.run()

* 這裏需要說明的是generate_annotation_json.py是爲了得到圖片標註的label_dict。通過這個代碼塊可以獲得我們需要的圖片標註字典,key是圖片具體地址, value是圖片的類別,具體實例如下:

{
"/images/hangs/862e67a8-5bd9-41f1-8c6d-876a3cb270df.JPG": 6, 
"/images/tags/adc264af-a76b-4477-9573-ac6c435decab.JPG": 3, 
"/images/tags/fd231f5a-b42c-43ba-9e9d-4abfbaf38853.JPG": 3, 
"/images/hangs/2e47d877-1954-40d6-bfa2-1b8e3952ebf9.jpg": 6, 
"/images/tileds/a07beddc-4b39-4865-8ee2-017e6c257e92.png": 5,
 "/images/models/642015c8-f29d-4930-b1a9-564f858c40e5.png": 4
}
  1. 如何運行代碼

(1)首先我們的文件夾構成形式是如下結構,其中images_root是圖片根文件夾,combinations, details, sizes, tags, models, tileds, hangs分別存放不同類別的圖片文件夾。

-<images_root>
   -<combinations>
      -圖片.jpg
   -<details>
      -圖片.jpg
   -<sizes>
      -圖片.jpg
   -<tags>
      -圖片.jpg
   -<models>
      -圖片.jpg
   -<tileds>
      -圖片.jpg
   -<hangs>
      -圖片.jpg

(2)建立文件夾TFRecord,並將generate_tfrecord.pygenerate_annotation_json.py這兩個python文件放入文件夾內,需要注意的是我們需要將 generate_tfrecord.py文件中字典word2number_dict換成自己的字典(即key是放不同類別的圖片文件夾名稱,value是對應的分類number)

    word2number_dict = { 
        "combinations": 0,
        "details": 1,
        "sizes": 2,
        "tags": 3,
        "models": 4,
        "tileds": 5,
        "hangs": 6
    }

(3)直接執行代碼 python3/python2 ./TFRecord/generate_tfrecord.py --image_dir="images_root地址" --record_path="你想要保存record地址(.record文件全路徑)"即可。如下是一個實例:

python3 generate_tfrecord.py --image_dir /images/ --record_path /classify/data/train_tfrecord/train.record



 

TFRecord讀取


上面我們介紹瞭如何生成TFRecord,現在我們嘗試如何通過使用隊列讀取讀取我們的TFRecord。
讀取TFRecord可以通過tensorflow兩個個重要的函數實現,分別是tf.train.string_input_producertf.TFRecordReadertf.parse_single_example解析器。如下圖

AnimatedFileQueues.gif

 

四、 讀取TFRecord的簡單實現方式

解析TFRecord有兩種解析方式一種是利用tf.parse_single_example, 另一種是通過tf.contrib.slim(* 推薦使用)。
1. 第一種方式(tf.parse_single_example)解析步驟如下
(1).第一步,我們將train.record文件讀入到隊列中,如下所示:
filename_queue = tf.train.string_input_producer([tfrecords_filename])
(2) 第二步,我們需要通過TFRecord將生成的隊列讀入

reader = tf.TFRecordReader()
 _, serialized_example = reader.read(filename_queue) #返回文件名和文件

(3)第三步, 通過解析器tf.parse_single_example將我們的example解析出來。
 

  1. 第二種方式(tf.contrib.slim)解析步驟如下

(1) 第一步, 我們要設置decoder = slim.tfexample_decoder.TFExampleDecoder(keys_to_features, items_to_handlers), 其中key_to_features這個字典需要和TFrecord文件中定義的字典項匹配,items_to_handlers中的關鍵字可以是任意值,但是它的handler的初始化參數必須要來自於keys_to_features中的關鍵字。

(2) 第二步, 我們要設定dataset = slim.dataset.Dataset(params), 其中params包括:
a. data_source: 爲tfrecord文件地址
b. reader: 一般設置爲tf.TFRecordReader閱讀器
c. decoder: 爲第一步設置的decoder
d. num_samples: 樣本數量
e. items_to_description: 對樣本及標籤的描述
f. num_classes: 分類的數量

(3) 第三步, 我們設置provider = slim.dataset_data_provider.DatasetDataProvider(params), 其中params包括 :
a. dataset: 第二步驟我們生成的數據集
b. num_reader: 並行閱讀器數量
c. shuffle: 是否打亂
d. num_epochs:每個數據源被讀取的次數,如果設爲None數據將會被無限循環的讀取
e. common_queue_capacity:讀取數據隊列的容量,默認爲256
f. scope:範圍
g. common_queue_min:讀取數據隊列的最小容量。

(4) 第四步, 我們可以通過provider.get得到我們需要的數據了。

3. 對不同圖片大小的TFRecord讀取並resize成相同大小
reshape_same_size函數來對圖片進行resize,這樣我們可以對我們的圖片進行batch操作了,因爲有的神經網絡訓練需要一個batch一個batch操作,不同大小的圖片在組成一個batch的時候會報錯,因此我們我通過後期處理可以更好的對圖片進行batch操作。
或者直接通過resized_image = tf.squeeze(tf.image.resize_bilinear([image], size=[FLAG.resize_height, FLAG.resize_width]))即可。

五、tf.contrib.slim模塊讀取TFrecord文件完整代碼實例

# -*- coding: utf-8 -*-
# @Time    : 2018/12/1 11:06
# @Author  : MaochengHu
# @Email   : [email protected]
# @File    : read_tfrecord.py
# @Software: PyCharm
import os
import tensorflow as tf

flags = tf.app.flags
flags.DEFINE_string('tfrecord_path', '/data1/humaoc_file/classify/data/train_tfrecord/train.record', 'path to tfrecord file')
flags.DEFINE_integer('resize_height', 800, 'resize height of image')
flags.DEFINE_integer('resize_width', 800, 'resize width of image')
FLAG = flags.FLAGS
slim = tf.contrib.slim


def print_data(image, resized_image, label, height, width):
    with tf.Session() as sess:
        init_op = tf.global_variables_initializer()
        sess.run(init_op)
        coord = tf.train.Coordinator()
        threads = tf.train.start_queue_runners(coord=coord)
        for i in range(10):
            print("______________________image({})___________________".format(i))
            print_image, print_resized_image, print_label, print_height, print_width = sess.run([image, resized_image, label, height, width])
            print("resized_image shape is: ", print_resized_image.shape)
            print("image shape is: ", print_image.shape)
            print("image label is: ", print_label)
            print("image height is: ", print_height)
            print("image width is: ", print_width)
        coord.request_stop()
        coord.join(threads)

def reshape_same_size(image, output_height, output_width):
    """Resize images by fixed sides.
    
    Args:
        image: A 3-D image `Tensor`.
        output_height: The height of the image after preprocessing.
        output_width: The width of the image after preprocessing.

    Returns:
        resized_image: A 3-D tensor containing the resized image.
    """
    output_height = tf.convert_to_tensor(output_height, dtype=tf.int32)
    output_width = tf.convert_to_tensor(output_width, dtype=tf.int32)

    image = tf.expand_dims(image, 0)
    resized_image = tf.image.resize_nearest_neighbor(
        image, [output_height, output_width], align_corners=False)
    resized_image = tf.squeeze(resized_image)
    return resized_image


def read_tfrecord(tfrecord_path, num_samples=14635, num_classes=7, resize_height=800, resize_width=800):
    keys_to_features = {
        'image/encoded': tf.FixedLenFeature([], default_value='', dtype=tf.string,),
        'image/format': tf.FixedLenFeature([], default_value='jpeg', dtype=tf.string),
        'image/class/label': tf.FixedLenFeature([], tf.int64, default_value=0),
        'image/height': tf.FixedLenFeature([], tf.int64, default_value=0),
        'image/width': tf.FixedLenFeature([], tf.int64, default_value=0)
    }

    items_to_handlers = {
        'image': slim.tfexample_decoder.Image(image_key='image/encoded', format_key='image/format', channels=3),
        'label': slim.tfexample_decoder.Tensor('image/class/label', shape=[]),
        'height': slim.tfexample_decoder.Tensor('image/height', shape=[]),
        'width': slim.tfexample_decoder.Tensor('image/width', shape=[])
    }
    decoder = slim.tfexample_decoder.TFExampleDecoder(keys_to_features, items_to_handlers)

    labels_to_names = None
    items_to_descriptions = {
        'image': 'An image with shape image_shape.',
        'label': 'A single integer between 0 and 9.'}

    dataset = slim.dataset.Dataset(
        data_sources=tfrecord_path,
        reader=tf.TFRecordReader,
        decoder=decoder,
        num_samples=num_samples,
        items_to_descriptions=None,
        num_classes=num_classes,
    )

    provider = slim.dataset_data_provider.DatasetDataProvider(dataset=dataset,
                                                              num_readers=3,
                                                              shuffle=True,
                                                              common_queue_capacity=256,
                                                              common_queue_min=128,
                                                              seed=None)
    image, label, height, width = provider.get(['image', 'label', 'height', 'width'])
    resized_image = tf.squeeze(tf.image.resize_bilinear([image], size=[resize_height, resize_width]))
    return resized_image, label, image, height, width




def main():
    resized_image, label, image, height, width = read_tfrecord(tfrecord_path=FLAG.tfrecord_path,
                                                               resize_height=FLAG.resize_height,
                                                               resize_width=FLAG.resize_width)
    #resized_image = reshape_same_size(image, FLAG.resize_height, FLAG.resize_width)
    #resized_image = tf.squeeze(tf.image.resize_bilinear([image], size=[FLAG.resize_height, FLAG.resize_width]))
    print_data(image, resized_image, label, height, width)
  


if __name__ == '__main__':
    main()

代碼運行方式

python3 read_tfrecord.py --tfrecord_path /data1/humaoc_file/classify/data/train_tfrecord/train.record --resize_height 800 --resize_width 800

 

 

最終我們可以看到我們讀取文件的部分內容:

______________________image(0)___________________
resized_image shape is:  (800, 800, 3)
image shape is:  (2000, 1333, 3)
image label is:  5
image height is:  2000
image width is:  1333
______________________image(1)___________________
resized_image shape is:  (800, 800, 3)
image shape is:  (667, 1000, 3)
image label is:  0
image height is:  667
image width is:  1000
______________________image(2)___________________
resized_image shape is:  (800, 800, 3)
image shape is:  (667, 1000, 3)
image label is:  3
image height is:  667
image width is:  1000
______________________image(3)___________________
resized_image shape is:  (800, 800, 3)
image shape is:  (800, 800, 3)
image label is:  5
image height is:  800
image width is:  800
______________________image(4)___________________
resized_image shape is:  (800, 800, 3)
image shape is:  (1424, 750, 3)
image label is:  0
image height is:  1424
image width is:  750
______________________image(5)___________________
resized_image shape is:  (800, 800, 3)
image shape is:  (1196, 1000, 3)
image label is:  6
image height is:  1196
image width is:  1000
______________________image(6)___________________
resized_image shape is:  (800, 800, 3)
image shape is:  (667, 1000, 3)
image label is:  5
image height is:  667
image width is:  1000

轉載地址:https://www.jianshu.com/p/b480e5fcb638

 

 

發表評論
所有評論
還沒有人評論,想成為第一個評論的人麼? 請在上方評論欄輸入並且點擊發布.
相關文章