import matplotlib.pyplot as plt
import numpy as np
import os
import PIL
import tensorflow as tf

from tensorflow import keras
from tensorflow.keras import layers
from tensorflow.keras.models import Sequential

"""flower_photo/
  daisy/
  dandelion/
  roses/
  sunflowers/
  tulips/"""
import pathlib

dataset_url = "https://storage.googleapis.com/download.tensorflow.org/example_images/flower_photos.tgz"
data_dir = tf.keras.utils.get_file('flower_photos', origin=dataset_url, untar=True)
print(data_dir)
print(type(data_dir))
data_dir = pathlib.Path(data_dir)
print(data_dir)
print(type(data_dir))

image_count = len(list(data_dir.glob('*/*.jpg')))
print(image_count)
roses = list(data_dir.glob('roses/*'))
img0 = PIL.Image.open(str(roses[0]))
plt.imshow(img0)
plt.show()

batch_size = 32
img_height = 180
img_width = 180

# It's good practice to use a validation split when developing your model.
# Let's use 80% of the images for training, and 20% for validation.
train_ds = tf.keras.preprocessing.image_dataset_from_directory(
    data_dir,
    validation_split=0.2,
    subset="training",
    seed=123,
    image_size=(img_height, img_width),
    batch_size=batch_size)

val_ds = tf.keras.preprocessing.image_dataset_from_directory(
    data_dir,
    validation_split=0.2,
    subset="validation",
    seed=123,
    image_size=(img_height, img_width),
    batch_size=batch_size)

class_names = train_ds.class_names
print(class_names)

# 图片可视化
plt.figure(figsize=(10, 10))
for images, labels in train_ds.take(1):
    for i in range(30):
        ax = plt.subplot(3, 10, i + 1)
        plt.imshow(images[i].numpy().astype("uint8"))
        plt.title(class_names[labels[i]])
        plt.axis("off")
plt.show()

for image_batch, labels_batch in train_ds:
    print(image_batch.shape)
    print(labels_batch.shape)
    break

AUTOTUNE = tf.data.AUTOTUNE

train_ds = train_ds.cache().shuffle(1000).prefetch(buffer_size=AUTOTUNE)
val_ds = val_ds.cache().prefetch(buffer_size=AUTOTUNE)

normalization_layer = layers.experimental.preprocessing.Rescaling(1. / 255)
normalized_ds = train_ds.map(lambda x, y: (normalization_layer(x), y))
image_batch, labels_batch = next(iter(normalized_ds))
first_image = image_batch[0]
# Notice the pixels values are now in `[0,1]`.
print(np.min(first_image), np.max(first_image))

data_augmentation = keras.Sequential(
    [
        layers.experimental.preprocessing.RandomFlip("horizontal",
                                                     input_shape=(img_height,
                                                                  img_width,
                                                                  3)),
        layers.experimental.preprocessing.RandomRotation(0.1),
        layers.experimental.preprocessing.RandomZoom(0.1),
    ]
)
plt.figure(figsize=(10, 10))
for images, _ in train_ds.take(1):
    for i in range(9):
        augmented_images = data_augmentation(images)
        ax = plt.subplot(3, 3, i + 1)
        plt.imshow(augmented_images[0].numpy().astype("uint8"))
        plt.axis("off")

num_classes = 5
model = Sequential([
    data_augmentation,
    layers.experimental.preprocessing.Rescaling(1. / 255),
    layers.Conv2D(16, 3, padding='same', activation='relu'),
    layers.MaxPooling2D(),
    layers.Conv2D(32, 3, padding='same', activation='relu'),
    layers.MaxPooling2D(),
    layers.Conv2D(64, 3, padding='same', activation='relu'),
    layers.MaxPooling2D(),
    layers.Conv2D(128, 3, padding='same', activation='relu'),
    layers.MaxPooling2D(),
    layers.Dropout(0.15),
    layers.Flatten(),
    layers.Dense(128, activation='relu'),
    layers.Dense(num_classes)
])

model.compile(optimizer='adam',
              loss=tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True),
              metrics=['accuracy'])
model.summary()

epochs = 15
history = model.fit(
    train_ds,
    validation_data=val_ds,
    epochs=epochs
)

model.save("./model/Flowers_1227.h5")   #保存模型

#读取并调用模型
pre_model = tf.keras.models.load_model("./model/Flowers_1227.h5")


acc = history.history['accuracy']
val_acc = history.history['val_accuracy']

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

epochs_range = range(epochs)

plt.figure(figsize=(8, 8))
plt.subplot(1, 2, 1)
plt.plot(epochs_range, acc, label='Training Accuracy')
plt.plot(epochs_range, val_acc, label='Validation Accuracy')
plt.legend(loc='lower right')
plt.title('Training and Validation Accuracy')

plt.subplot(1, 2, 2)
plt.plot(epochs_range, loss, label='Training Loss')
plt.plot(epochs_range, val_loss, label='Validation Loss')
plt.legend(loc='upper right')
plt.title('Training and Validation Loss')
plt.show()

sunflower_url = "https://storage.googleapis.com/download.tensorflow.org/example_images/592px-Red_sunflower.jpg"
sunflower_path = tf.keras.utils.get_file('Red_sunflower', origin=sunflower_url)

img = keras.preprocessing.image.load_img(
    sunflower_path, target_size=(img_height, img_width)
)
img_array = keras.preprocessing.image.img_to_array(img)
img_array = tf.expand_dims(img_array, 0)  # Create a batch

predictions = pre_model.predict(img_array)
score = tf.nn.softmax(predictions[0])

print(
    "This image most likely belongs to  with a :.2f percent confidence."
        .format(class_names[np.argmax(score)], 100 * np.max(score))
)

修改的代码包含：（1）修改了模型，增加了一个卷积层；（2）增加模型保存代码；（3）增加模型读取代码，并使用读取到的模型预测图片

1.1 模型运行参数总结

Model: "sequential_1"
_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
sequential (Sequential)      (None, 180, 180, 3)       0         
_________________________________________________________________
rescaling_1 (Rescaling)      (None, 180, 180, 3)       0         
_________________________________________________________________
conv2d (Conv2D)              (None, 180, 180, 16)      448       
_________________________________________________________________
max_pooling2d (MaxPooling2D) (None, 90, 90, 16)        0         
_________________________________________________________________
conv2d_1 (Conv2D)            (None, 90, 90, 32)        4640      
_________________________________________________________________
max_pooling2d_1 (MaxPooling2 (None, 45, 45, 32)        0         
_________________________________________________________________
conv2d_2 (Conv2D)            (None, 45, 45, 64)        18496     
_________________________________________________________________
max_pooling2d_2 (MaxPooling2 (None, 22, 22, 64)        0         
_________________________________________________________________
conv2d_3 (Conv2D)            (None, 22, 22, 128)       73856     
_________________________________________________________________
max_pooling2d_3 (MaxPooling2 (None, 11, 11, 128)       0         
_________________________________________________________________
dropout (Dropout)            (None, 11, 11, 128)       0         
_________________________________________________________________
flatten (Flatten)            (None, 15488)             0         
_________________________________________________________________
dense (Dense)                (None, 128)               1982592   
_________________________________________________________________
dense_1 (Dense)              (None, 5)                 645       
=================================================================
Total params: 2,080,677
Trainable params: 2,080,677
Non-trainable params: 0

1.2模型训练效果

15次epoch有75.61%的精度，增加训练次数应该还有一定提升空间。

Epoch 1/15
92/92 [==============================] - 99s 1s/step - loss: 1.3126 - accuracy: 0.4087 - val_loss: 1.0708 - val_accuracy: 0.5477
Epoch 2/15
92/92 [==============================] - 88s 957ms/step - loss: 1.0561 - accuracy: 0.5562 - val_loss: 0.9844 - val_accuracy: 0.5872
Epoch 3/15
92/92 [==============================] - 89s 966ms/step - loss: 0.9517 - accuracy: 0.6117 - val_loss: 1.0068 - val_accuracy: 0.6035
Epoch 4/15
92/92 [==============================] - 84s 913ms/step - loss: 0.8743 - accuracy: 0.6550 - val_loss: 0.8538 - val_accuracy: 0.6580
Epoch 5/15
92/92 [==============================] - 82s 891ms/step - loss: 0.8065 - accuracy: 0.6809 - val_loss: 0.8371 - val_accuracy: 0.6703
Epoch 6/15
92/92 [==============================] - 82s 892ms/step - loss: 0.7623 - accuracy: 0.7115 - val_loss: 0.8203 - val_accuracy: 0.7016
Epoch 7/15
92/92 [==============================] - 94s 1s/step - loss: 0.7309 - accuracy: 0.7245 - val_loss: 0.7539 - val_accuracy: 0.7057
Epoch 8/15
92/92 [==============================] - 90s 982ms/step - loss: 0.6928 - accuracy: 0.7262 - val_loss: 0.7811 - val_accuracy: 0.7166
Epoch 9/15
92/92 [==============================] - 88s 955ms/step - loss: 0.6840 - accuracy: 0.7333 - val_loss: 0.8314 - val_accuracy: 0.6703
Epoch 10/15
92/92 [==============================] - 81s 877ms/step - loss: 0.6591 - accuracy: 0.7565 - val_loss: 0.7585 - val_accuracy: 0.7153
Epoch 11/15
92/92 [==============================] - 83s 899ms/step - loss: 0.6195 - accuracy: 0.7633 - val_loss: 0.7600 - val_accuracy: 0.7125
Epoch 12/15
92/92 [==============================] - 86s 934ms/step - loss: 0.6006 - accuracy: 0.7657 - val_loss: 0.6871 - val_accuracy: 0.7262
Epoch 13/15
92/92 [==============================] - 86s 934ms/step - loss: 0.5736 - accuracy: 0.7762 - val_loss: 0.6955 - val_accuracy: 0.7452
Epoch 14/15
92/92 [==============================] - 82s 897ms/step - loss: 0.5523 - accuracy: 0.7871 - val_loss: 0.7513 - val_accuracy: 0.7234
Epoch 15/15
92/92 [==============================] - 86s 935ms/step - loss: 0.5379 - accuracy: 0.7956 - val_loss: 0.6591 - val_accuracy: 0.7561

图片数据增强后的效果图：

2.模型的保存

训练模型的保存实际上只需一行代码就行，在模型训练完成之后，我们将模型保存到指定的路径并给模型命名。模型保存的格式是.h5后缀的格式，这种文件是hdf5格式的数据，我们可以使用专门的软件打开查看模型相关参数。

在model.fit()训练完模型之后，保存模型到model文件夹下：

model.save("./model/Flowers_1227.h5")   #保存模型

运行完成之后在项目文件下可以看到model文件夹，文件中可以看到我们保存的模型：

模型大小有45.7M.

3.读取模型model

读取代码也只需要一行，如下：

#读取并调用模型
pre_model = tf.keras.models.load_model("./model/Flowers_1227.h5")

4.使用模型进行图片预测

根据上面读取到的模型直接进行图片预测。

可以省去前面的数据训练部分，直接使用后面的部分代码，读取模型然后进行图片预测。

继续运行上文中的代码后面部分，即最后面的部分是预测一张图片属于什么类型的。

运行结果是：

This image most likely belongs to sunflowers with a 98.13 percent confidence.

读取模型进行预测的代码如下：

import matplotlib.pyplot as plt
import numpy as np
import os
import PIL
import tensorflow as tf

from tensorflow import keras
from tensorflow.keras import layers
from tensorflow.keras.models import Sequential

"""flower_photo/
  daisy/
  dandelion/
  roses/
  sunflowers/
  tulips/"""
import pathlib

batch_size = 32
img_height = 180
img_width = 180

dataset_url = "https://storage.googleapis.com/download.tensorflow.org/example_images/flower_photos.tgz"
data_dir = tf.keras.utils.get_file('flower_photos', origin=dataset_url, untar=True)
train_ds = tf.keras.preprocessing.image_dataset_from_directory(
    data_dir,
    validation_split=0.2,
    subset="training",
    seed=123,
    image_size=(img_height, img_width),
    batch_size=batch_size)
class_names = train_ds.class_names
print(class_names)

#读取并调用模型
pre_model = tf.keras.models.load_model("./model/Flowers_1227.h5")

sunflower_url = "https://storage.googleapis.com/download.tensorflow.org/example_images/592px-Red_sunflower.jpg"
sunflower_path = tf.keras.utils.get_file('Red_sunflower', origin=sunflower_url)

img = keras.preprocessing.image.load_img(
    sunflower_path, target_size=(img_height, img_width)
)
img_array = keras.preprocessing.image.img_to_array(img)
img_array = tf.expand_dims(img_array, 0)  # Create a batch

predictions = pre_model.predict(img_array)
score = tf.nn.softmax(predictions[0])
print(score)

print(
    "This image most likely belongs to  with a :.2f percent confidence."
        .format(class_names[np.argmax(score)], 100 * np.max(score))
)

运行结果：

Found 3670 files belonging to 5 classes.
Using 2936 files for training.
2022-12-27 22:34:13.075000: I tensorflow/core/platform/cpu_feature_guard.cc:142] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX AVX2
To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
['daisy', 'dandelion', 'roses', 'sunflowers', 'tulips']
2022-12-27 22:34:14.205000: I tensorflow/compiler/mlir/mlir_graph_optimization_pass.cc:176] None of the MLIR Optimization Passes are enabled (registered 2)
tf.Tensor([1.1012822e-04 5.2587932e-04 4.4165729e-03 9.8130155e-01 1.3645952e-02], shape=(5,), dtype=float32)
This image most likely belongs to sunflowers with a 98.13 percent confidence.

我们可以看到预测结果中，score = tf.nn.softmax(predictions[0])，代表的是该图片属于每种类型的概率大小，第四个是最大的，即第四个对应的是sunflowers类别，因此可以认为预测的结果就是sunflowers。

这里预测的图片是使用在线下载的一张图片进行预测的，实际上我们可以读取我们本地路径下的文件进行大批量的预测，并将每张图片预测结果保存到文本文件中用于后续的分析。