基于python3.6+tensorflow2.2的猫狗分类案例

unzip_save.py：数据集解压+存储

import os
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2'  # 不显示等级2以下的提示信息
import zipfile
# import random
# import tensorflow as tf
# from tensorflow.keras.optimizers import RMSprop
# from tensorflow.keras.preprocessing.image import ImageDataGenerator
# from shutil import copyfile


local_zip = 'tmp/cats-and-dogs.zip' # 数据集压缩包路径
zip_ref = zipfile.ZipFile(local_zip, 'r') # 打开压缩包，以读取方式
zip_ref.extractall('E:/Python/pythonProject_1/tmp') # 解压到以下路径
zip_ref.close()

print(len(os.listdir('tmp/cats-and-dogs/PetImages/Cat/'))) # 读取出解压文件夹“Cat”文件夹下面文件数量
print(len(os.listdir('tmp/cats-and-dogs/PetImages/Dog/'))) # 读取出解压文件夹“Dog”文件夹下面文件数量

try:
    os.mkdir('/tmp/cats-and-dogs') # 创建虚拟路径名称'/tmp/cats-and-dogs'
    os.mkdir('/tmp/cats-and-dogs/training') # 创建虚拟路径名称'/tmp/cats-and-dogs/training'
    os.mkdir('/tmp/cats-and-dogs/testing') # 创建虚拟路径名称'/tmp/cats-and-dogs/testing'
    os.mkdir('/tmp/cats-and-dogs/training/cats') # 创建虚拟路径名称'/tmp/cats-and-dogs/training/cats'
    os.mkdir('/tmp/cats-and-dogs/training/dogs') # 创建虚拟路径名称'/tmp/cats-and-dogs/training/dogs'
    os.mkdir('/tmp/cats-and-dogs/testing/cats') # 创建虚拟路径名称'/tmp/cats-and-dogs/testing/cats'
    os.mkdir('/tmp/cats-and-dogs/testing/dogs') # 创建虚拟路径名称'/tmp/cats-and-dogs/testing/dogs'
except  OSError:
    pass

model_training_fit.py：模型构建+训练+评估

model_training_fit流程图：

import os
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2'  # 不显示等级2以下的提示信息
# import shutil
# import random
import tensorflow as tf
# from tensorflow import keras
from tensorflow.keras.optimizers import RMSprop
from tensorflow.keras.preprocessing.image import ImageDataGenerator
# from shutil import copyfile
# import matplotlib.image as mpimg
import matplotlib.pyplot as plt


#======== 数据预处理 =========
TRAINING_DIR = "E:/Python/pythonProject_1/cats_vs_dogs/tmp/cats-and-dogs/training/"
train_datagen = ImageDataGenerator(rescale = 1/255) # 归一化，减少计算量
train_generator = train_datagen.flow_from_directory(TRAINING_DIR, batch_size = 100, class_mode = 'binary', target_size = (150,150))

VALIDATION_DIR = "E:/Python/pythonProject_1/cats_vs_dogs/tmp/cats-and-dogs/testing/"
validation_datagen = ImageDataGenerator(rescale = 1/255)
validation_generator = validation_datagen.flow_from_directory(VALIDATION_DIR, batch_size = 100, class_mode = 'binary', target_size = (150,150))

# Expected Output :
# Found 22498 images belonging to 2 classes.
# Found 2500 images belonging to 2 classes.

#======== 模型构建 =========
model = tf.keras.models.Sequential([
    tf.keras.layers.Conv2D(16, (3, 3), activation = 'relu', input_shape = (150, 150, 3)), # 输入参数：过滤器数量，过滤器尺寸，激活函数：relu， 输入图像尺寸
    tf.keras.layers.MaxPooling2D(2, 2), # 池化：增强特征
    tf.keras.layers.Conv2D(32, (3, 3), activation = 'relu'), # 输入参数：过滤器数量、过滤器尺寸、激活函数：relu
    tf.keras.layers.MaxPooling2D(2, 2),
    tf.keras.layers.Conv2D(64, (3, 3), activation = 'relu'),
    tf.keras.layers.MaxPooling2D(2, 2),
    tf.keras.layers.Flatten(), # 输入层
    tf.keras.layers.Dense(512, activation = 'relu'), # 神经元数量：512 ，激活函数：relu
    tf.keras.layers.Dense(1, activation = 'sigmoid')
])
#======== 模型参数编译 =========
model.compile(optimizer = RMSprop(lr = 0.001), loss = 'binary_crossentropy', metrics = ['accuracy'])
# 输入参数：lr:学习步长   RMSprop算法：修正学习步长  loss：损失函数  metrics：网络性能度量

#======== 模型训练 =========
# Note that this may take some time.
history = model.fit_generator(train_generator, epochs = 2, verbose = 1, validation_data = validation_generator) # verbose=1 表示记录网络拟合过程

model.save('E:/Python/pythonProject_1/cats_vs_dogs/model.h5') # model 保存

#-----------------------------------------------------------
# Retrieve a list of list result on training and test data
# set for each training epoch
#-----------------------------------------------------------
acc = history.history['accuracy']
val_acc = history.history['val_accuracy']
loss = history.history['loss']
val_loss = history.history['val_loss']

epochs = range(len(acc)) # Get number of epochs

#-----------------------------------------------------------
# Plot training and validation accuracy per epoch
#-----------------------------------------------------------
plt.plot(epochs, acc, 'r', "train_acc")
plt.plot(epochs ,val_acc, 'b', "valida_acc")
plt.title("training and validation accuracy")
plt.figure()
plt.show()
# 曲线呈直线是因为epochs/轮次太少
#-----------------------------------------------------------
# Plot training and validation loss per epoch
#-----------------------------------------------------------
plt.plot(epochs, loss, 'r', "train_loss")
plt.plot(epochs ,val_loss, 'b', "valida_loss")
plt.title("training and validation loss")
plt.figure()
plt.show()
# 曲线呈直线是因为epochs/轮次太少

main.py：模型预测

# Here's a codeblock just for fun,You should be able to upload an image here
# and have it claassified without crashing

import numpy as np
# from google.colab import files
from tensorflow.keras.preprocessing import image
from tensorflow import keras
model = keras.models.load_model('model.h5')

# upload = files.upload()
#
# for fn in upload.key():
#     # predicting images
#     path = '/content/' + fn
#     img = image.load_img(path, target_size=(150, 150))
#     x = image.img_to_array(img)
#     x = np.expand_dims(x, axis=0)
#
#     images = np.vstack([x])
#     classes = model.predict(images, batch_size=10)
#     print(classes[0])
#     if classes[0]>0.5:
#         print(fn + " is a dog")
#     else:
#         print(fn + " is a cat")


# predicting images
path = 'E:\Python\pythonProject_1\cats_vs_dogs\dog.jpg'
img = image.load_img(path, target_size=(150, 150))
x = image.img_to_array(img)
x = np.expand_dims(x, axis=0)

images = np.vstack([x])
classes = model.predict(images, batch_size=10)
print(classes[0])
if classes[0]>0.5:
    print(" is a dog")
else:
    print(" is a cat")

---

评估结果：

Epoch 1/2
 36/225 [===>..........................] - ETA: 2:39 - loss: 0.9044 - accuracy: 0.5269E:\pycharm\Anaconda\envs\pythonProject_1\lib\site-packages\PIL\TiffImagePlugin.py:793: UserWarning: Truncated File Read
  warnings.warn(str(msg))
225/225 [==============================] - 194s 864ms/step - loss: 0.6645 - accuracy: 0.6406 - val_loss: 0.6875 - val_accuracy: 0.6060
Epoch 2/2
225/225 [==============================] - 192s 855ms/step - loss: 0.5199 - accuracy: 0.7426 - val_loss: 0.4600 - val_accuracy: 0.7936

---

预测结果：

path = 'E:\Python\pythonProject_1\cat.jpg'
img = image.load_img(path, target_size=(150, 150))
x = image.img_to_array(img)
x = np.expand_dims(x, axis=0)
images = np.vstack([x])
classes = model.predict(images, batch_size=10)
print(classes[0])
if classes[0]>0.5:
    print(" is a dog")
else:
    print(" is a cat")
    
[0.]
 is a cat

path = 'E:\Python\pythonProject_1\dog.jpg'
img = image.load_img(path, target_size=(150, 150))
x = image.img_to_array(img)
x = np.expand_dims(x, axis=0)
images = np.vstack([x])
classes = model.predict(images, batch_size=10)
print(classes[0])
if classes[0]>0.5:
    print(" is a dog")
else:
    print(" is a cat")
    
[1.]
 is a dog

---

备注：

数据预处理等效图：

---

卷积、池化等效流程图(颜色和像素值仅作参考，与实际值无关)：

---

模型结构：

model.summary()
Model: "sequential"
_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
conv2d (Conv2D)              (None, 148, 148, 16)      448       
_________________________________________________________________
max_pooling2d (MaxPooling2D) (None, 74, 74, 16)        0         
_________________________________________________________________
conv2d_1 (Conv2D)            (None, 72, 72, 32)        4640      
_________________________________________________________________
max_pooling2d_1 (MaxPooling2 (None, 36, 36, 32)        0         
_________________________________________________________________
conv2d_2 (Conv2D)            (None, 34, 34, 64)        18496     
_________________________________________________________________
max_pooling2d_2 (MaxPooling2 (None, 17, 17, 64)        0         
_________________________________________________________________
flatten (Flatten)            (None, 18496)             0         
_________________________________________________________________
dense (Dense)                (None, 512)               9470464   
_________________________________________________________________
dense_1 (Dense)              (None, 1)                 513       
=================================================================
Total params: 9,494,561
Trainable params: 9,494,561
Non-trainable params: 0
_________________________________________________________________