ImageDataGenerator是keras中帮助我们实现对图片数据的读取和处理的API。
keras.preprocessing.image.ImageDataGenerator.flow_from_directory
具体参数见keras官方文档:
https://keras-cn.readthedocs.io/en/latest/
一个读取实例:
train_datagen = ImageDataGenerator(rescale=1./255,
shear_range=0.2,
zoom_range=0.2,
horizontal_flip=True)
test_datagen = ImageDataGenerator(rescale=1./255)
#训练路径
train_generator = train_datagen.flow_from_directory(‘H:\train’,
target_size=(28,28),
batch_size=32,
)
#val路径
validation_generator = test_datagen.flow_from_directory(‘H:\val’,
target_size=(28,28),
batch_size=32,
)
model.fit_generator(train_generator,
steps_per_epoch=420,
epochs=50,
callbacks=callback_list,
validation_data=validation_generator,
validation_steps=25,
verbose=1)
#fit使用train_generator和validation_generator
文件夹路径结构:
dog 和pig 是要类别,里面存放img/png/…等常用格式的该类图片就行。val路径一样:
比如: