导入所需库

import numpy as np import matplotlib.pyplot as plt import tensorflow as tf from tensorflow.keras.preprocessing.image import ImageDataGenerator from tensorflow.keras.models import Sequential from tensorflow.keras.layers import Conv2D, MaxPooling2D, Flatten, Dense, Dropout from tensorflow.keras.optimizers import Adam from sklearn.preprocessing import LabelBinarizer

数据集收集、读取

data_dir = r"C:\Users\24423\Desktop\Course Design for Deep Learning\image" train_dir = os.path.join(data_dir, 'train') test_dir = os.path.join(data_dir, 'test')

加载数据集并进行数据可视化

batch_size = 32 train_datagen = ImageDataGenerator(rescale=1./255, shear_range=0.2, zoom_range=0.2, horizontal_flip=True) train_generator = train_datagen.flow_from_directory(train_dir, target_size=(224, 224), batch_size=batch_size, class_mode='categorical') test_datagen = ImageDataGenerator(rescale=1./255) test_generator = test_datagen.flow_from_directory(test_dir, target_size=(224, 224), batch_size=batch_size, class_mode='categorical')

将标签进行one-hot编码

label_binarizer = LabelBinarizer() train_labels = label_binarizer.fit_transform(train_generator.classes) test_labels = label_binarizer.fit_transform(test_generator.classes)

将数据集划分为训练集和验证集

train_size = train_generator.samples test_size = test_generator.samples train_steps = np.ceil(train_size / batch_size) test_steps = np.ceil(test_size / batch_size)

搭建卷积神经网络模型

model = Sequential() model.add(Conv2D(32, (3, 3), activation='relu', input_shape=(224, 224, 3))) model.add(MaxPooling2D((2, 2))) model.add(Conv2D(64, (3, 3), activation='relu')) model.add(MaxPooling2D((2, 2))) model.add(Conv2D(128, (3, 3), activation='relu')) model.add(MaxPooling2D((2, 2))) model.add(Conv2D(256, (3, 3), activation='relu')) model.add(MaxPooling2D((2, 2))) model.add(Flatten()) model.add(Dense(512, activation='relu')) model.add(Dropout(0.5)) model.add(Dense(256, activation='relu')) model.add(Dropout(0.5)) model.add(Dense(3, activation='softmax'))

参数优化

lr = 0.0001 epochs = 20 optimizer = Adam(lr) model.compile(loss='categorical_crossentropy', optimizer=optimizer, metrics=['accuracy'])

训练模型

history = model.fit(train_generator, steps_per_epoch=train_steps, epochs=epochs, validation_data=test_generator, validation_steps=test_steps)

显示模型结构

model.summary()

绘制准确率及Loss曲线图

plt.plot(history.history['accuracy'], label='accuracy') plt.plot(history.history['val_accuracy'], label='val_accuracy') plt.plot(history.history['loss'], label='loss') plt.plot(history.history['val_loss'], label='val_loss') plt.legend() plt.show()

图像识别展示

test_images, test_labels = next(test_generator) predictions = model.predict(test_images) for i in range(batch_size): image = test_images[i] true_label = test_labels[i] predicted_label = predictions[i] plt.imshow(image) plt.title("True: {0}, Predicted: {1}".format(np.argmax(true_label), np.argmax(predicted_label))) plt.show(