使用 TensorFlow Keras 对灰度图像进行 CNN 训练

问题描述

我使用

将RGB彩色图像数据集转换为灰度图像

from PIL import Image,ImageEnhance
    im = Image.open('../DR datasets/DR-60.20.20/val/level_4/4.png')
    im=im.convert('L')
    enhancer1 = ImageEnhance.Contrast(im)
    enhancer2 = ImageEnhance.Sharpness(im)
    factorBrightness =100
    factorSharpness=2
    output= enhancer1.enhance(factorBrightness)
    output= enhancer2.enhance(factorSharpness)
    print("Brightness and Sharpness Changed")
    output.save('../DR datasets/greyImage/4/greyimage.png')

然后我将此数据集输入到 CNN 模型中

model = models.Sequential()
model.add(layers.Conv2D(32,3,padding="same",kernel_initializer='he_uniform',activation= 'relu',input_shape=(224,224,1)))
model.add(layers.MaxPooling2D((2,2),padding="same"))
model.add(layers.Conv2D(32,activation= 'relu'))
model.add(layers.MaxPooling2D((2,padding="same"))
model.add(layers.Conv2D(64,activation='relu'))
model.add(layers.MaxPooling2D((2,padding="same"))
model.add(layers.Conv2D(128,padding="same"))
model.add(layers.Flatten())
model.add(layers.Dense(100,activation='relu'))
model.add(layers.Dense(50,activation='relu'))
model.add(layers.Dense(25,activation='relu'))
model.add(layers.Dense(15,activation='relu'))
model.add(layers.Dropout(0.5))
model.add(layers.Dense(5,activation="softmax"))
img_height,img_width = 224,224
# the no. imgaes to load at each iteration
batch_size = 32
# only rescaling
train_datagen =  ImageDataGenerator(
    rescale=1./255
)
test_datagen =  ImageDataGenerator(
    rescale=1./255
)
# these are generators for train/test data that will read pictures #found in the defined subfolders of 'data/'
print('Total number of images for "training":')
train_generator = train_datagen.flow_from_directory(
train_data_dir,target_size = (img_height,img_width),batch_size = batch_size,class_mode = "categorical",shuffle = False,color_mode='grayscale'
)

#train_generator = train_generator.reshape(x_train.shape[0],x_train.shape[1],x_train.shape[2],1)
print('Total number of images for "validation":')
val_generator = test_datagen.flow_from_directory(
val_data_dir,color_mode='grayscale',shuffle=False)
print('Total number of images for "testing":')
test_generator = test_datagen.flow_from_directory(
test_data_dir,shuffle=False)

然后我使用以下代码编译和训练模型：

learning_rate =1e-4
epochs = 30
checkpoint = ModelCheckpoint("mqmodifiedvgg_classifier.h5",monitor = 'val_acc',verbose=1,save_best_only=True,save_weights_only=False,mode='auto',period=1)
model.compile(loss="categorical_crossentropy",optimizer=tensorflow.optimizers.Adam(lr=learning_rate),metrics = ['acc']) 
history = model.fit(train_generator,epochs=epochs,shuffle=True,validation_data=val_generator,#steps_per_epoch=100,callbacks=[checkpoint]
                              )

但是它在 1 个 epoch 完成后没有训练模型它会抛出错误

NotFoundError：没有算法起作用！ [[节点 sequence_5/conv2d_25/Conv2D（定义于 :4)]] [操作：__inference_test_function_22208]

函数调用栈：test_function

我使用 len(image.shape) 检查长度，它返回 2，这意味着图像是 1 个通道，但仍然没有让模型训练提前致谢

解决方法

暂无找到可以解决该程序问题的有效方法，小编努力寻找整理中！

如果你已经找到好的解决方法，欢迎将解决方案带上本链接一起发送给小编。

小编邮箱:dio#foxmail.com (将#修改为@）

conv-neural-network grayscale keras keras python tensorflow tensorflow tensorflow