-- coding: utf-8 --

'使用VGG16预训练模型进行图像特征提取，并构建全连接神经网络进行分类'

import os import matplotlib.pyplot as plt import numpy as np from keras.applications import VGG16 from keras.preprocessing.image import ImageDataGenerator

data_dir = 'ss'+os.sep+'image' train_dir = os.path.join(data_dir, 'train') test_dir = os.path.join(data_dir, 'test')

vgg_model = VGG16(weights='imagenet', include_top=False, input_shape=(150,150,3)) vgg_model.summary()

datagen = ImageDataGenerator(rescale=1./255)

batch_size = 20

#定义提取特征函数 def extract_features(directory, sample_count): #VGG16最后池化层的输出特征初始化储存张量 features = np.zeros(shape=(sample_count,4,4,512)) labels = np.zeros(shape=(sample_count)) #从数据样本文件夹按batch_size数加载图像 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: features_batch = vgg_model.predict(inputs_batch) features[ibatch_size : (i+1) * batch_size] = features_batch labels[ibatch_size : (i+1) * batch_size] = labels_batch i += 1 # if i *batch_size >= sample_count: break return features, labels

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)

train_features = np.reshape(train_features, (2000, 44512)) #validation_features = np.reshape(validation_features, (1000, 44512)) test_features = np.reshape(test_features, (1000, 44512))

from keras import models from keras import layers from keras import optimizers

model = models.Sequential() model.add(layers.Dense(256,activation='relu',input_dim=44512)) model.add(layers.Dropout(0.5)) model.add(layers.Dense(1,activation='sigmoid')) model.compile(optimizer=optimizers.Adam(lr=2e-5), loss='binary_crossentropy', metrics=['acc'])

history = model.fit(train_features,train_labels, epochs=30, batch_size=batch_size, validation_data=(test_features, test_labels))

acc = history.history['acc'] val_acc = history.history['val_acc'] epochs = range(1, len(acc) + 1) plt.figure() plt.plot(epochs,acc,'bo',label='训练集准确率') plt.plot(epochs,val_acc,'b',label='验证集准确率') plt.title('训练集与验证集准确率曲线') plt.legend() plt.show()

acc = history.history['loss'] val_acc = history.history['val_loss'] epochs = range(1,len(acc) + 1) plt.figure() plt.plot(epochs,acc,'bo',label='训练集损失') plt.plot(epochs,val_acc,'b',label='验证集损失') plt.title('训练集与验证集损失曲线') plt.legend() plt.show()