卷积神经网络的猫狗识别——基于Tensorflow和Keras

一、环境配置

1. 安装Anaconda3
   详细安装配置:
   https://blog.csdn.net/cungudafa/article/details/104573389
2. 配置TensorFlow、Keras
   ①创建虚拟环境
   输入下面命令

conda create -n tf1 python=3.6
#tf1是自己为创建虚拟环境取的名字，后面python的版本可以根据自己需求进行选择

②安装tensorflow和keras

pip install 包名
#直接这样安装可以由于网络的原因，安装失败或者安装很慢
#解决方式：
pip install -i https://pypi.tuna.tsinghua.edu.cn/simple 包名
#此次安装命令如下：
pip install -i https://pypi.tuna.tsinghua.edu.cn/simple tensorflow==1.14.0
pip install -i https://pypi.tuna.tsinghua.edu.cn/simple keras==2.2.5

二、数据集的准备

1. 数据集的下载
   kaggle网站的数据集下载地址：
   https://www.kaggle.com/lizhensheng/-2000
2. 数据集的分类
   将下载的数据集进行解压操作
   
   按照命名进行分类
   分类前
   
   分类后
   
3. 分类代码如下

import tensorflow as tf
import keras
import os, shutil 
# 原始目录所在的路径
original_dataset_dir = 'G:\\Cat_And_Dog\\kaggle\\train\\'

# 数据集分类后的目录
base_dir = 'G:\\Cat_And_Dog\\kaggle\\cats_and_dogs_small'
os.mkdir(base_dir)

# # 训练、验证、测试数据集的目录
train_dir = os.path.join(base_dir, 'train')
os.mkdir(train_dir)
validation_dir = os.path.join(base_dir, 'validation')
os.mkdir(validation_dir)
test_dir = os.path.join(base_dir, 'test')
os.mkdir(test_dir)

# 猫训练图片所在目录
train_cats_dir = os.path.join(train_dir, 'cats')
os.mkdir(train_cats_dir)

# 狗训练图片所在目录
train_dogs_dir = os.path.join(train_dir, 'dogs')
os.mkdir(train_dogs_dir)

# 猫验证图片所在目录
validation_cats_dir = os.path.join(validation_dir, 'cats')
os.mkdir(validation_cats_dir)

# 狗验证数据集所在目录
validation_dogs_dir = os.path.join(validation_dir, 'dogs')
os.mkdir(validation_dogs_dir)

# 猫测试数据集所在目录
test_cats_dir = os.path.join(test_dir, 'cats')
os.mkdir(test_cats_dir)

# 狗测试数据集所在目录
test_dogs_dir = os.path.join(test_dir, 'dogs')
os.mkdir(test_dogs_dir)

# 将前1000张猫图像复制到train_cats_dir
fnames = ['cat.{}.jpg'.format(i) for i in range(1000)]
for fname in fnames:
    src = os.path.join(original_dataset_dir, fname)
    dst = os.path.join(train_cats_dir, fname)
    shutil.copyfile(src, dst)

# 将下500张猫图像复制到validation_cats_dir
fnames = ['cat.{}.jpg'.format(i) for i in range(1000, 1500)]
for fname in fnames:
    src = os.path.join(original_dataset_dir, fname)
    dst = os.path.join(validation_cats_dir, fname)
    shutil.copyfile(src, dst)
    
# 将下500张猫图像复制到test_cats_dir
fnames = ['cat.{}.jpg'.format(i) for i in range(1500, 2000)]
for fname in fnames:
    src = os.path.join(original_dataset_dir, fname)
    dst = os.path.join(test_cats_dir, fname)
    shutil.copyfile(src, dst)
    
# 将前1000张狗图像复制到train_dogs_dir
fnames = ['dog.{}.jpg'.format(i) for i in range(1000)]
for fname in fnames:
    src = os.path.join(original_dataset_dir, fname)
    dst = os.path.join(train_dogs_dir, fname)
    shutil.copyfile(src, dst)
    
# 将下500张狗图像复制到validation_dogs_dir
fnames = ['dog.{}.jpg'.format(i) for i in range(1000, 1500)]
for fname in fnames:
    src = os.path.join(original_dataset_dir, fname)
    dst = os.path.join(validation_dogs_dir, fname)
    shutil.copyfile(src, dst)
    
# 将下500张狗图像复制到test_dogs_dir
fnames = ['dog.{}.jpg'.format(i) for i in range(1500, 2000)]
for fname in fnames:
    src = os.path.join(original_dataset_dir, fname)
    dst = os.path.join(test_dogs_dir, fname)
    shutil.copyfile(src, dst)

4. 查看分类后，统计对应目录下图片数量

#输出数据集对应目录下图片数量
print('total training cat images:', len(os.listdir(train_cats_dir)))
print('total training dog images:', len(os.listdir(train_dogs_dir)))
print('total validation cat images:', len(os.listdir(validation_cats_dir)))
print('total validation dog images:', len(os.listdir(validation_dogs_dir)))
print('total test cat images:', len(os.listdir(test_cats_dir)))
print('total test dog images:', len(os.listdir(test_dogs_dir)))

三、猫狗分类的实例——基准模型

1. 构建网络模型

#网络模型构建
from keras import layers
from keras import models
#keras的序贯模型
model = models.Sequential()
#卷积层，卷积核是3*3，激活函数relu
model.add(layers.Conv2D(32, (3, 3), activation='relu',
                        input_shape=(150, 150, 3)))
#最大池化层
model.add(layers.MaxPooling2D((2, 2)))
#卷积层，卷积核2*2，激活函数relu
model.add(layers.Conv2D(64, (3, 3), activation='relu'))
#最大池化层
model.add(layers.MaxPooling2D((2, 2)))
#卷积层，卷积核是3*3，激活函数relu
model.add(layers.Conv2D(128, (3, 3), activation='relu'))
#最大池化层
model.add(layers.MaxPooling2D((2, 2)))
#卷积层，卷积核是3*3，激活函数relu
model.add(layers.Conv2D(128, (3, 3), activation='relu'))
#最大池化层
model.add(layers.MaxPooling2D((2, 2)))
#flatten层，用于将多维的输入一维化，用于卷积层和全连接层的过渡
model.add(layers.Flatten())
#全连接，激活函数relu
model.add(layers.Dense(512, activation='relu'))
#全连接，激活函数sigmoid
model.add(layers.Dense(1, activation='sigmoid'))

查看模型各层的参数状况

#输出模型各层的参数状况
model.summary()

2. 配置训练方法

model.compile(optimizer = 优化器，
              loss = 损失函数，
              metrics = ["准确率”])

其中，优化器和损失函数可以是字符串形式的名字，也可以是函数形式。

from keras import optimizers

model.compile(loss='binary_crossentropy',
              optimizer=optimizers.RMSprop(lr=1e-4),
              metrics=['acc'])

3. 文件中图像转换成所需格式
   将训练和验证的图片，调整为150*150

from keras.preprocessing.image import ImageDataGenerator

# 所有图像将按1/255重新缩放
train_datagen = ImageDataGenerator(rescale=1./255)
test_datagen = ImageDataGenerator(rescale=1./255)

train_generator = train_datagen.flow_from_directory(
        # 这是目标目录
        train_dir,
        # 所有图像将调整为150x150
        target_size=(150, 150),
        batch_size=20,
        # 因为我们使用二元交叉熵损失，我们需要二元标签
        class_mode='binary')

validation_generator = test_datagen.flow_from_directory(
        validation_dir,
        target_size=(150, 150),
        batch_size=20,
        class_mode='binary')

查看处理结果

#查看上面对于图片预处理的处理结果
for data_batch, labels_batch in train_generator:
    print('data batch shape:', data_batch.shape)
    print('labels batch shape:', labels_batch.shape)
    break

4. 模型训练并保存生成的模型

#模型训练过程
history = model.fit_generator(
      train_generator,
      steps_per_epoch=100,
      epochs=30,
      validation_data=validation_generator,
      validation_steps=50)
#保存训练得到的的模型
model.save('G:\\Cat_And_Dog\\kaggle\\cats_and_dogs_small_1.h5')

5. 结果可视化

#对于模型进行评估，查看预测的准确性
import matplotlib.pyplot as plt

acc = history.history['acc']
val_acc = history.history['val_acc']
loss = history.history['loss']
val_loss = history.history['val_loss']

epochs = range(len(acc))

plt.plot(epochs, acc, 'bo', label='Training acc')
plt.plot(epochs, val_acc, 'b', label='Validation acc')
plt.title('Training and validation accuracy')
plt.legend()

plt.figure()

plt.plot(epochs, loss, 'bo', label='Training loss')
plt.plot(epochs, val_loss, 'b', label='Validation loss')
plt.title('Training and validation loss')
plt.legend()

plt.show()

由可视化结果，可以发现训练的loss是成上升趋势。所以，训练获得的模型存在一些问题，导致模型过拟合。过拟合是为了得到一致假设而使假设变得过度严格，实际训练得到的模型的分类效果不佳。

四、根据基准模型进行调整

为了解决过拟合问题，可以减小模型复杂度，也可以用一系列手段去对冲，比如增加数据（图像增强、人工合成或者多搜集真实数据）、L1/L2正则化、dropout正则化等。这里主要介绍CV中最常用的图像增强。

1. 图像增强
   利用图像生成器定义一些常见的图像变换，图像增强就是通过对于图像进行变换，从而，增强图像中的有用信息。

#该部分代码及以后的代码，用于替代基准模型中分类后面的代码（执行代码前，需要先将之前分类的目录删掉，重写生成分类，否则，会发生错误）
from keras.preprocessing.image import ImageDataGenerator
datagen = ImageDataGenerator(
      rotation_range=40,
      width_shift_range=0.2,
      height_shift_range=0.2,
      shear_range=0.2,
      zoom_range=0.2,
      horizontal_flip=True,
      fill_mode='nearest')

①rotation_range
一个角度值(0-180)，在这个范围内可以随机旋转图片
②width_shift和height_shift
范围(作为总宽度或高度的一部分)，在其中可以随机地垂直或水平地转换图片
③shear_range
用于随机应用剪切转换
④zoom_range
用于在图片内部随机缩放
⑤horizontal_flip
用于水平随机翻转一半的图像——当没有假设水平不对称时(例如真实世界的图片)
⑥fill_mode
用于填充新创建像素的策略，它可以在旋转或宽度/高度移动之后出现

2. 查看增强后的图像

import matplotlib.pyplot as plt
# This is module with image preprocessing utilities
from keras.preprocessing import image
fnames = [os.path.join(train_cats_dir, fname) for fname in os.listdir(train_cats_dir)]
# We pick one image to "augment"
img_path = fnames[3]
# Read the image and resize it
img = image.load_img(img_path, target_size=(150, 150))
# Convert it to a Numpy array with shape (150, 150, 3)
x = image.img_to_array(img)
# Reshape it to (1, 150, 150, 3)
x = x.reshape((1,) + x.shape)
# The .flow() command below generates batches of randomly transformed images.
# It will loop indefinitely, so we need to `break` the loop at some point!
i = 0
for batch in datagen.flow(x, batch_size=1):
    plt.figure(i)
    imgplot = plt.imshow(image.array_to_img(batch[0]))
    i += 1
    if i % 4 == 0:
        break
plt.show()

3. 网络模型增加一层dropout

#网络模型构建
from keras import layers
from keras import models
#keras的序贯模型
model = models.Sequential()
#卷积层，卷积核是3*3，激活函数relu
model.add(layers.Conv2D(32, (3, 3), activation='relu',
                        input_shape=(150, 150, 3)))
#最大池化层
model.add(layers.MaxPooling2D((2, 2)))
#卷积层，卷积核2*2，激活函数relu
model.add(layers.Conv2D(64, (3, 3), activation='relu'))
#最大池化层
model.add(layers.MaxPooling2D((2, 2)))
#卷积层，卷积核是3*3，激活函数relu
model.add(layers.Conv2D(128, (3, 3), activation='relu'))
#最大池化层
model.add(layers.MaxPooling2D((2, 2)))
#卷积层，卷积核是3*3，激活函数relu
model.add(layers.Conv2D(128, (3, 3), activation='relu'))
#最大池化层
model.add(layers.MaxPooling2D((2, 2)))
#flatten层，用于将多维的输入一维化，用于卷积层和全连接层的过渡
model.add(layers.Flatten())
#退出层
model.add(layers.Dropout(0.5))
#全连接，激活函数relu
model.add(layers.Dense(512, activation='relu'))
#全连接，激活函数sigmoid
model.add(layers.Dense(1, activation='sigmoid'))
#输出模型各层的参数状况
model.summary()
from keras import optimizers

model.compile(loss='binary_crossentropy',
              optimizer=optimizers.RMSprop(lr=1e-4),
              metrics=['acc'])

4. 训练模型

train_datagen = ImageDataGenerator(
    rescale=1./255,
    rotation_range=40,
    width_shift_range=0.2,
    height_shift_range=0.2,
    shear_range=0.2,
    zoom_range=0.2,
    horizontal_flip=True,)
# Note that the validation data should not be augmented!
test_datagen = ImageDataGenerator(rescale=1./255)
train_generator = train_datagen.flow_from_directory(
        # This is the target directory
        train_dir,
        # All images will be resized to 150x150
        target_size=(150, 150),
        batch_size=32,
        # Since we use binary_crossentropy loss, we need binary labels
        class_mode='binary')
validation_generator = test_datagen.flow_from_directory(
        validation_dir,
        target_size=(150, 150),
        batch_size=32,
        class_mode='binary')
history = model.fit_generator(
      train_generator,
      steps_per_epoch=100,
      epochs=100,
      validation_data=validation_generator,
      validation_steps=50)
model.save('G:\\Cat_And_Dog\\kaggle\\cats_and_dogs_small_2.h5')

5. 结果可视化

acc = history.history['acc']
val_acc = history.history['val_acc']
loss = history.history['loss']
val_loss = history.history['val_loss']
epochs = range(len(acc))
plt.plot(epochs, acc, 'bo', label='Training acc')
plt.plot(epochs, val_acc, 'b', label='Validation acc')
plt.title('Training and validation accuracy')
plt.legend()
plt.figure()
plt.plot(epochs, loss, 'bo', label='Training loss')
plt.plot(epochs, val_loss, 'b', label='Validation loss')
plt.title('Training and validation loss')
plt.legend()
plt.show()

只进行数据增强的可视化结果


数据增强和dropout层增加的可视化结果

对比基准模型来看，可以很清楚的发现loss的整体趋势是变小的。对比，只进行图像增强获得的模型和进行图像增强与添加dropout层获得的模型，可以发现前者在训练过程中波动会更大，后者在准确上小于前者。两者虽然在准确率有所变小，但是都避免了过拟合

五、使用VGG19实现猫狗分类

1. 初始化一个VGG19网络实例

from keras.applications import VGG19
conv_base = VGG19(weights = 'imagenet',include_top = False,input_shape=(150, 150, 3))
conv_base.summary()

首次运行时候，会自动从对应网站下载h5格式文件

上面下载很慢，而且还有可能在中途挂掉，因此建议将网址复制到浏览器上，直接下载。然后，将下载的文件，放到对应的目录下
我下载存放的位置

其模型网络结构

2. 将猫狗数据集传递给神经网络
   将分类后的猫狗数据集传递给神经网络，让它把图片的隐含信息给抽取出来

import os 
import numpy as np
from keras.preprocessing.image import ImageDataGenerator
# 数据集分类后的目录
base_dir = 'E:\\Cat_And_Dog\\kaggle\\cats_and_dogs_small'
train_dir = os.path.join(base_dir, 'train')
validation_dir = os.path.join(base_dir, 'validation')
test_dir = os.path.join(base_dir, 'test')
datagen = ImageDataGenerator(rescale = 1. / 255)
batch_size = 20
def extract_features(directory, sample_count):
    features = np.zeros(shape = (sample_count, 4, 4, 512))
    labels = np.zeros(shape = (sample_count))
    generator = datagen.flow_from_directory(directory, target_size = (150, 150), 
                                            batch_size = batch_size,
                                            class_mode = 'binary')
    i = 0
    for inputs_batch, labels_batch in generator:
        #把图片输入VGG16卷积层，让它把图片信息抽取出来
        features_batch = conv_base.predict(inputs_batch)
        #feature_batch 是 4*4*512结构
        features[i * batch_size : (i + 1)*batch_size] = features_batch
        labels[i * batch_size : (i+1)*batch_size] = labels_batch
        i += 1
        if i * batch_size >= sample_count :
            #for in 在generator上的循环是无止境的，因此我们必须主动break掉
            break
        return features , labels
#extract_features 返回数据格式为(samples, 4, 4, 512)
train_features, train_labels = extract_features(train_dir, 2000)
validation_features, validation_labels = extract_features(validation_dir, 1000)
test_features, test_labels = extract_features(test_dir, 1000)	

3. 将抽取的特征输入到我们自己的神经层中进行分类训练

from keras import models
from keras import layers
from keras import optimizers
#构造我们自己的网络层对输出数据进行分类
model = models.Sequential()
model.add(layers.Dense(256, activation='relu', input_dim = 4 * 4 * 512))
model.add(layers.Dropout(0.5))
model.add(layers.Dense(1, activation = 'sigmoid'))
model.compile(optimizer=optimizers.RMSprop(lr = 2e-5), loss = 'binary_crossentropy', metrics = ['acc'])
history = model.fit(train_features, train_labels, epochs = 30, batch_size = 20, 
                    validation_data = (validation_features, validation_labels))

4. 训练结果和校验结果的可视化

import matplotlib.pyplot as plt
acc = history.history['acc']
val_acc = history.history['val_acc']
loss = history.history['loss']
val_loss = history.history['val_loss']
epochs = range(1, len(acc) + 1)
plt.plot(epochs, acc, 'bo', label = 'Train_acc')
plt.plot(epochs, val_acc, 'b', label = 'Validation acc')
plt.title('Trainning and validation accuracy')
plt.legend()
plt.figure()
plt.plot(epochs, loss, 'bo', label = 'Training loss')
plt.plot(epochs, val_loss, 'b', label = 'Validation loss')
plt.title('Training and validation loss')
plt.legend()
plt.show()

六、参考链接

1.【TensorFlow&Keras】入门猫狗数据集实验–理解卷积神经网络CNN
2.基于Tensorflow和Keras实现卷积神经网络CNN
3.基于jupyter notebook的python编程-----猫狗数据集的阶段分类得到模型精度并进行数据集优化