TensorFLow 不同大小图片的TFrecords存取

最新推荐文章于 2022-11-03 12:17:31 发布

Wayne2019

最新推荐文章于 2022-11-03 12:17:31 发布

阅读量3.8k

点赞数 1

分类专栏： TensorFlow 文章标签： tensorflow tfrecords 深度学习与计算机视觉

本文链接：https://blog.csdn.net/Wayne2019/article/details/77894794

版权

TensorFlow 专栏收录该内容

22 篇文章 0 订阅

订阅专栏

刚搬到美国，终于有时间搞竞赛了。。

全部存入一个TFrecords文件，然后读取并显示第一张。
不多写了，直接贴代码。

from PIL import Image
import numpy as np
import matplotlib.pyplot as plt
import tensorflow as tf


IMAGE_PATH = 'test/'
tfrecord_file = IMAGE_PATH + 'test.tfrecord'
writer = tf.python_io.TFRecordWriter(tfrecord_file)


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 get_image_binary(filename):
    """ You can read in the image using tensorflow too, but it's a drag
        since you have to create graphs. It's much easier using Pillow and NumPy
    """
    image = Image.open(filename)
    image = np.asarray(image, np.uint8)
    shape = np.array(image.shape, np.int32)
    return shape, image.tobytes() # convert image to raw data bytes in the array.

def write_to_tfrecord(label, shape, binary_image, tfrecord_file):
    """ This example is to write a sample to TFRecord file. If you want to write
    more samples, just use a loop.
    """
    # write label, shape, and image content to the TFRecord file
    example = tf.train.Example(features=tf.train.Features(feature={
                'label': _int64_feature(label),
                'h': _int64_feature(shape[0]),
                'w': _int64_feature(shape[1]),
                'c': _int64_feature(shape[2]),
                'image': _bytes_feature(binary_image)
                }))
    writer.write(example.SerializeToString())


def write_tfrecord(label, image_file, tfrecord_file):
    shape, binary_image = get_image_binary(image_file)
    write_to_tfrecord(label, shape, binary_image, tfrecord_file)
   # print(shape)



def main():
    # assume the image has the label Chihuahua, which corresponds to class number 1
    label = [1,2]
    image_files = [IMAGE_PATH + 'a.jpg', IMAGE_PATH + 'b.jpg']

    for i in range(2):
        write_tfrecord(label[i], image_files[i], tfrecord_file)
    writer.close()

    batch_size = 2

    filename_queue = tf.train.string_input_producer([tfrecord_file])  
    reader = tf.TFRecordReader()  
    _, serialized_example = reader.read(filename_queue)  

    img_features = tf.parse_single_example(  
                                        serialized_example,  
                                        features={  
                                               'label': tf.FixedLenFeature([], tf.int64),  
                                               'h': tf.FixedLenFeature([], tf.int64),
                                               'w': tf.FixedLenFeature([], tf.int64),
                                               'c': tf.FixedLenFeature([], tf.int64),
                                               'image': tf.FixedLenFeature([], tf.string),  
                                               })  

    h = tf.cast(img_features['h'], tf.int32)
    w = tf.cast(img_features['w'], tf.int32)
    c = tf.cast(img_features['c'], tf.int32)

    image = tf.decode_raw(img_features['image'], tf.uint8)  
    image = tf.reshape(image, [h, w, c])

    label = tf.cast(img_features['label'],tf.int32) 
    label = tf.reshape(label, [1])

  #  image = tf.image.resize_images(image, (500,500))
    #image, label = tf.train.batch([image, label],  batch_size= batch_size)  


    with tf.Session() as sess:
        coord = tf.train.Coordinator()
        threads = tf.train.start_queue_runners(coord=coord)
        image, label=sess.run([image, label])
        coord.request_stop()
        coord.join(threads)

        print(label)

        plt.figure()
        plt.imshow(image)
        plt.show()


if __name__ == '__main__':
    main()

全部存入一个TFrecords文件，然后按照batch_size读取，注意需要将图片变成一样大才能按照batch_size读取。

from PIL import Image
import numpy as np
import matplotlib.pyplot as plt
import tensorflow as tf


IMAGE_PATH = 'test/'
tfrecord_file = IMAGE_PATH + 'test.tfrecord'
writer = tf.python_io.TFRecordWriter(tfrecord_file)


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 get_image_binary(filename):
    """ You can read in the image using tensorflow too, but it's a drag
        since you have to create graphs. It's much easier using Pillow and NumPy
    """
    image = Image.open(filename)
    image = np.asarray(image, np.uint8)
    shape = np.array(image.shape, np.int32)
    return shape, image.tobytes() # convert image to raw data bytes in the array.

def write_to_tfrecord(label, shape, binary_image, tfrecord_file):
    """ This example is to write a sample to TFRecord file. If you want to write
    more samples, just use a loop.
    """
    # write label, shape, and image content to the TFRecord file
    example = tf.train.Example(features=tf.train.Features(feature={
                'label': _int64_feature(label),
                'h': _int64_feature(shape[0]),
                'w': _int64_feature(shape[1]),
                'c': _int64_feature(shape[2]),
                'image': _bytes_feature(binary_image)
                }))
    writer.write(example.SerializeToString())


def write_tfrecord(label, image_file, tfrecord_file):
    shape, binary_image = get_image_binary(image_file)
    write_to_tfrecord(label, shape, binary_image, tfrecord_file)
   # print(shape)



def main():
    # assume the image has the label Chihuahua, which corresponds to class number 1
    label = [1,2]
    image_files = [IMAGE_PATH + 'a.jpg', IMAGE_PATH + 'b.jpg']

    for i in range(2):
        write_tfrecord(label[i], image_files[i], tfrecord_file)
    writer.close()

    batch_size = 2

    filename_queue = tf.train.string_input_producer([tfrecord_file])  
    reader = tf.TFRecordReader()  
    _, serialized_example = reader.read(filename_queue)  

    img_features = tf.parse_single_example(  
                                        serialized_example,  
                                        features={  
                                               'label': tf.FixedLenFeature([], tf.int64),  
                                               'h': tf.FixedLenFeature([], tf.int64),
                                               'w': tf.FixedLenFeature([], tf.int64),
                                               'c': tf.FixedLenFeature([], tf.int64),
                                               'image': tf.FixedLenFeature([], tf.string),  
                                               })  

    h = tf.cast(img_features['h'], tf.int32)
    w = tf.cast(img_features['w'], tf.int32)
    c = tf.cast(img_features['c'], tf.int32)

    image = tf.decode_raw(img_features['image'], tf.uint8)  
    image = tf.reshape(image, [h, w, c])

    label = tf.cast(img_features['label'],tf.int32) 
    label = tf.reshape(label, [1])

    image = tf.image.resize_images(image, (224,224))
    image = tf.reshape(image, [224, 224, 3])
    image, label = tf.train.batch([image, label],  batch_size= batch_size)  


    with tf.Session() as sess:
        coord = tf.train.Coordinator()
        threads = tf.train.start_queue_runners(coord=coord)
        image, label=sess.run([image, label])
        coord.request_stop()
        coord.join(threads)

        print(image.shape)
        print(label)

        plt.figure()
        plt.imshow(image[0,:,:,0])
        plt.show()

        plt.figure()
        plt.imshow(image[0,:,:,1])
        plt.show()

        image1 = image[0,:,:,:]
        print(image1.shape)
        print(image1.dtype)
        im = Image.fromarray(np.uint8(image1)) #参考numpy和图片的互转：http://blog.csdn.net/zywvvd/article/details/72810360
        im.show()

if __name__ == '__main__':
    main()

输出是

(2, 224, 224, 3)
[[1]
 [2]]
第一张图片的三种显示（略）

封装成函数：

# -*- coding: utf-8 -*-
"""
Created on Fri Sep  8 14:38:15 2017

@author: wayne


"""


'''
本文参考了以下代码，在多个不同大小图片存取方面做了重新开发：
https://github.com/chiphuyen/stanford-tensorflow-tutorials/blob/master/examples/09_tfrecord_example.py
http://blog.csdn.net/hjxu2016/article/details/76165559
https://stackoverflow.com/questions/41921746/tensorflow-varlenfeature-vs-fixedlenfeature
https://github.com/tensorflow/tensorflow/issues/10492

后续：
-存入多个TFrecords文件的例子见
http://blog.csdn.net/xierhacker/article/details/72357651
-如何作shuffle和数据增强
string_input_producer (需要理解tf的数据流，标签队列的工作方式等等)
http://blog.csdn.net/liuchonge/article/details/73649251
'''

from PIL import Image
import numpy as np
import matplotlib.pyplot as plt
import tensorflow as tf


IMAGE_PATH = 'test/'
tfrecord_file = IMAGE_PATH + 'test.tfrecord'
writer = tf.python_io.TFRecordWriter(tfrecord_file)


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 get_image_binary(filename):
    """ You can read in the image using tensorflow too, but it's a drag
        since you have to create graphs. It's much easier using Pillow and NumPy
    """
    image = Image.open(filename)
    image = np.asarray(image, np.uint8)
    shape = np.array(image.shape, np.int32)
    return shape, image.tobytes() # convert image to raw data bytes in the array.

def write_to_tfrecord(label, shape, binary_image, tfrecord_file):
    """ This example is to write a sample to TFRecord file. If you want to write
    more samples, just use a loop.
    """
    # write label, shape, and image content to the TFRecord file
    example = tf.train.Example(features=tf.train.Features(feature={
                'label': _int64_feature(label),
                'h': _int64_feature(shape[0]),
                'w': _int64_feature(shape[1]),
                'c': _int64_feature(shape[2]),
                'image': _bytes_feature(binary_image)
                }))
    writer.write(example.SerializeToString())


def write_tfrecord(label, image_file, tfrecord_file):
    shape, binary_image = get_image_binary(image_file)
    write_to_tfrecord(label, shape, binary_image, tfrecord_file)


def read_and_decode(tfrecords_file, batch_size):  
    '''''read and decode tfrecord file, generate (image, label) batches 
    Args: 
        tfrecords_file: the directory of tfrecord file 
        batch_size: number of images in each batch 
    Returns: 
        image: 4D tensor - [batch_size, width, height, channel] 
        label: 1D tensor - [batch_size] 
    '''  
    # make an input queue from the tfrecord file  

    filename_queue = tf.train.string_input_producer([tfrecord_file])  
    reader = tf.TFRecordReader()  
    _, serialized_example = reader.read(filename_queue)  

    img_features = tf.parse_single_example(  
                                        serialized_example,  
                                        features={  
                                               'label': tf.FixedLenFeature([], tf.int64),  
                                               'h': tf.FixedLenFeature([], tf.int64),
                                               'w': tf.FixedLenFeature([], tf.int64),
                                               'c': tf.FixedLenFeature([], tf.int64),
                                               'image': tf.FixedLenFeature([], tf.string),  
                                               })  

    h = tf.cast(img_features['h'], tf.int32)
    w = tf.cast(img_features['w'], tf.int32)
    c = tf.cast(img_features['c'], tf.int32)

    image = tf.decode_raw(img_features['image'], tf.uint8)  
    image = tf.reshape(image, [h, w, c])

    label = tf.cast(img_features['label'],tf.int32) 
    label = tf.reshape(label, [1])

    ##########################################################  
    # you can put data augmentation here   
#    distorted_image = tf.random_crop(images, [530, 530, img_channel])
#    distorted_image = tf.image.random_flip_left_right(distorted_image)
#    distorted_image = tf.image.random_brightness(distorted_image, max_delta=63)
#    distorted_image = tf.image.random_contrast(distorted_image, lower=0.2, upper=1.8)
#    distorted_image = tf.image.resize_images(distorted_image, (imagesize,imagesize))
#    float_image = tf.image.per_image_standardization(distorted_image)

    image = tf.image.resize_images(image, (224,224))
    image = tf.reshape(image, [224, 224, 3])
    #image, label = tf.train.batch([image, label],  batch_size= batch_size)  

    image_batch, label_batch = tf.train.batch([image, label],  
                                                batch_size= batch_size,  
                                                num_threads= 64,   
                                                capacity = 2000)  
    return image_batch, tf.reshape(label_batch, [batch_size])  

def read_tfrecord2(tfrecord_file, batch_size):
    train_batch, train_label_batch = read_and_decode(tfrecord_file, batch_size)

    with tf.Session() as sess:
        coord = tf.train.Coordinator()
        threads = tf.train.start_queue_runners(coord=coord)
        train_batch, train_label_batch = sess.run([train_batch, train_label_batch])
        coord.request_stop()
        coord.join(threads)
    return train_batch, train_label_batch


def main():
    # assume the image has the label Chihuahua, which corresponds to class number 1
    label = [1,2]
    image_files = [IMAGE_PATH + 'a.jpg', IMAGE_PATH + 'b.jpg']

    for i in range(2):
        write_tfrecord(label[i], image_files[i], tfrecord_file)
    writer.close()

    batch_size = 2
   # read_tfrecord(tfrecord_file) # 读取一个图
    train_batch, train_label_batch = read_tfrecord2(tfrecord_file, batch_size)

    print(train_batch.shape)
    print(train_label_batch)

    plt.figure()
    plt.imshow(train_batch[0,:,:,0])
    plt.show()

    plt.figure()
    plt.imshow(train_batch[0,:,:,1])
    plt.show()

    train_batch1 = train_batch[0,:,:,:]
    print(train_batch.shape)
    print(train_batch1.dtype)
    im = Image.fromarray(np.uint8(train_batch1)) #参考numpy和图片的互转：http://blog.csdn.net/zywvvd/article/details/72810360
    im.show()

if __name__ == '__main__':
    main()