from keras.applications.vgg16 import VGG16#直接導(dǎo)入已經(jīng)訓(xùn)練好的VGG16網(wǎng)絡(luò)
from keras.preprocessing.image import load_img#load_image作用是載入圖片
from keras.preprocessing.image import img_to_array
from keras.applications.vgg16 import preprocess_input
from keras.applications.vgg16 import decode_predictions
 
model = VGG16()
image = load_img('D:\\photo\\dog.jpg',target_size=(224,224))#參數(shù)target_size用于設(shè)置目標(biāo)的大小，如此一來無論載入的原圖像大小如何，都會被標(biāo)準(zhǔn)化成統(tǒng)一的大小，這樣做是為了向神經(jīng)網(wǎng)絡(luò)中方便地輸入數(shù)據(jù)所需的。
image = img_to_array(image)#函數(shù)img_to_array會把圖像中的像素數(shù)據(jù)轉(zhuǎn)化成NumPy中的array，這樣數(shù)據(jù)才可以被Keras所使用。
#神經(jīng)網(wǎng)絡(luò)接收一張或多張圖像作為輸入，也就是說，輸入的array需要有4個維度： samples, rows, columns, and channels。由于我們僅有一個 sample（即一張image），我們需要對這個array進行reshape操作。
image = image.reshape((1,image.shape[0],image.shape[1],image.shape[2]))
image = preprocess_input(image)#對圖像進行預(yù)處理
y = model.predict(image)#預(yù)測圖像的類別
label = decode_predictions(y)#Keras提供了一個函數(shù)decode_predictions()，用以對已經(jīng)得到的預(yù)測向量進行解讀。該函數(shù)返回一個類別列表，以及類別中每個類別的預(yù)測概率，
label = label[0][0]
print('%s(%.2f%%)'%(label[1],label[2]*100))
# print(model.summary())

from keras.models import Sequential
from keras.layers.core import Flatten,Dense,Dropout
from keras.layers.convolutional import Convolution2D,MaxPooling2D,ZeroPadding2D
from keras.optimizers import SGD
import numpy as np
 
from keras.preprocessing import image
from keras.applications.imagenet_utils import preprocess_input, decode_predictions
import time
from keras import backend as K
K.set_image_dim_ordering('th')
def VGG_16(weights_path=None):
  model = Sequential()
 
  model.add(ZeroPadding2D((1, 1), input_shape=(3, 224, 224)))
  model.add(Convolution2D(64, (3, 3), activation='relu'))
  model.add(ZeroPadding2D((1, 1)))
  model.add(Convolution2D(64, (3, 3), activation='relu'))
  model.add(MaxPooling2D((2, 2), strides=(2, 2)))
 
  model.add(ZeroPadding2D((1, 1)))
  model.add(Convolution2D(128, (3, 3), activation='relu'))
  model.add(ZeroPadding2D((1, 1)))
  model.add(Convolution2D(128, (3, 3), activation='relu'))
  model.add(MaxPooling2D((2, 2), strides=(2, 2)))
 
  model.add(ZeroPadding2D((1, 1)))
  model.add(Convolution2D(256, (3, 3), activation='relu'))
  model.add(ZeroPadding2D((1, 1)))
  model.add(Convolution2D(256, (3, 3), activation='relu'))
  model.add(ZeroPadding2D((1, 1)))
  model.add(Convolution2D(256, (3, 3), activation='relu'))
  model.add(MaxPooling2D((2, 2), strides=(2, 2)))
 
  model.add(ZeroPadding2D((1, 1)))
  model.add(Convolution2D(512, (3, 3), activation='relu'))
  model.add(ZeroPadding2D((1, 1)))
  model.add(Convolution2D(512, (3, 3), activation='relu'))
  model.add(ZeroPadding2D((1, 1)))
  model.add(Convolution2D(512, (3, 3), activation='relu'))
  model.add(MaxPooling2D((2, 2), strides=(2, 2)))
 
  model.add(ZeroPadding2D((1, 1)))
  model.add(Convolution2D(512, (3, 3), activation='relu'))
  model.add(ZeroPadding2D((1, 1)))
  model.add(Convolution2D(512, (3, 3), activation='relu'))
  model.add(ZeroPadding2D((1, 1)))
  model.add(Convolution2D(512, (3, 3), activation='relu'))
  model.add(MaxPooling2D((2, 2), strides=(2, 2)))
 
  model.add(Flatten())
  model.add(Dense(4096, activation='relu'))
  model.add(Dropout(0.5))
  model.add(Dense(4096, activation='relu'))
  model.add(Dropout(0.5))
  model.add(Dense(1000, activation='softmax'))
 
  if weights_path:
    model.load_weights(weights_path,by_name=True)
 
  return model
 
model = VGG_16(weights_path='F:\\Kaggle\\vgg16_weights.h5')
sgd = SGD(lr=0.1, decay=1e-6, momentum=0.9, nesterov=True)
model.compile(optimizer=sgd, loss='categorical_crossentropy')
 
t0 = time.time()
img = image.load_img('D:\\photo\\dog.jpg', target_size=(224, 224))
x = image.img_to_array(img) # 三維（224，224，3）
x = np.expand_dims(x, axis=0) # 四維（1，224，224，3）#因為keras要求的維度是這樣的，所以要增加一個維度
x = preprocess_input(x) # 預(yù)處理
print(x.shape)
y_pred = model.predict(x) # 預(yù)測概率
 
t1 = time.time()
 
print("測試圖：", decode_predictions(y_pred)) # 輸出五個最高概率(類名, 語義概念, 預(yù)測概率)
print("耗時：", str((t1 - t0) * 1000), "ms")

# 使用VGG16模型
from keras.applications.vgg16 import VGG16
print('Start build VGG16 -------')
 
# 獲取vgg16的卷積部分，如果要獲取整個vgg16網(wǎng)絡(luò)需要設(shè)置:include_top=True
model_vgg16_conv = VGG16(weights='imagenet', include_top=False)
model_vgg16_conv.summary()
 
# 創(chuàng)建自己的輸入格式
# if K.image_data_format() == 'channels_first':
#  input_shape = (3, img_width, img_height)
# else:
#  input_shape = (img_width, img_height, 3)
 
input = Input(input_shape, name = 'image_input') # 注意，Keras有個層就是Input層
 
# 將vgg16模型原始輸入轉(zhuǎn)換成自己的輸入
output_vgg16_conv = model_vgg16_conv(input)
 
# output_vgg16_conv是包含了vgg16的卷積層，下面我需要做二分類任務(wù)，所以需要添加自己的全連接層
x = Flatten(name='flatten')(output_vgg16_conv)
x = Dense(4096, activation='relu', name='fc1')(x)
x = Dense(512, activation='relu', name='fc2')(x)
x = Dense(128, activation='relu', name='fc3')(x)
x = Dense(1, activation='softmax', name='predictions')(x)
 
# 最終創(chuàng)建出自己的vgg16模型
my_model = Model(input=input, output=x)
 
# 下面的模型輸出中，vgg16的層和參數(shù)不會顯示出，但是這些參數(shù)在訓(xùn)練的時候會更改
print('\nThis is my vgg16 model for the task')
my_model.summary()