Keras框架下的U-Net实现（可实现多类分割+代码）

U-Net网络结构

U-Net是一种全卷积网络，当卷积操作采用"padding = same"的形式，卷积不会使图片尺寸变小；maxpooling操作以2为步长，得到的结果是原图的$\frac{1}{4}$（宽和高均为原图的$\frac{1}{2}$）；后面有对应的上采样层恢复图像的尺寸。所以，最终可以得到和原图大小一样的分割结果。原文地址

U-Net的Keras实现

具体代码详见Github

Readme

可以在configuration.txt中修改适合自己数据的参数，如路径（数据路径，模型保存路径）、图像参数（h，w，channels，classes）、网络参数（N_epoch，batchsize）等。

部分代码

#!/usr/bin/env python
# -*- coding:utf-8 -*-
# Author:CaoZhihui
...
# --Build a net work-- #
def get_net():
    inputs = Input(shape=(N_channels, img_h, img_w))
    # Block 1
    conv1 = Conv2D(32, (3, 3), activation='relu', padding='same', kernel_initializer='he_normal')(inputs)
    conv1 = Conv2D(32, (3, 3), activation='relu', padding='same', kernel_initializer='he_normal')(conv1)
    pool1 = MaxPooling2D((2, 2), strides=(2, 2), name='block1_pool')(conv1)

    # Block 2
    conv2 = Conv2D(64, (3, 3), activation='relu', padding='same', kernel_initializer='he_normal')(pool1)
    conv2 = Conv2D(64, (3, 3), activation='relu', padding='same', kernel_initializer='he_normal')(conv2)
    pool2 = MaxPooling2D((2, 2), strides=(2, 2), name='block2_pool')(conv2)

    # Block 3
    conv3 = Conv2D(128, (3, 3), activation='relu', padding='same', kernel_initializer='he_normal')(pool2)
    conv3 = Conv2D(128, (3, 3), activation='relu', padding='same', kernel_initializer='he_normal')(conv3)
    pool3 = MaxPooling2D((2, 2), strides=(2, 2), name='block3_pool')(conv3)

    # Block 4
    conv4 = Conv2D(256, (3, 3), activation='relu', padding='same', kernel_initializer='he_normal')(pool3)
    conv4 = Conv2D(256, (3, 3), activation='relu', padding='same', kernel_initializer='he_normal')(conv4)
    pool4 = MaxPooling2D((2, 2), strides=(2, 2), name='block4_pool')(conv4)

    # Block 5
    conv5 = Conv2D(512, (3, 3), activation='relu', padding='same', kernel_initializer='he_normal')(pool4)
    conv5 = Conv2D(512, (3, 3), activation='relu', padding='same', kernel_initializer='he_normal')(conv5)

    up6 = merge([UpSampling2D(size=(2, 2))(conv5), conv4], mode='concat', concat_axis=1)
    conv6 = Conv2D(256, (3, 3), activation='relu', padding='same', kernel_initializer='he_normal')(up6)
    conv6 = Conv2D(256, (3, 3), activation='relu', padding='same', kernel_initializer='he_normal')(conv6)

    up7 = merge([UpSampling2D(size=(2, 2))(conv6), conv3], mode='concat', concat_axis=1)
    conv7 = Conv2D(128, (3, 3), activation='relu', padding='same', kernel_initializer='he_normal')(up7)
    conv7 = Conv2D(128, (3, 3), activation='relu', padding='same', kernel_initializer='he_normal')(conv7)

    up8 = merge([UpSampling2D(size=(2, 2))(conv7), conv2], mode='concat', concat_axis=1)
    conv8 = Conv2D(64, (3, 3), activation='relu', padding='same', kernel_initializer='he_normal')(up8)
    conv8 = Conv2D(64, (3, 3), activation='relu', padding='same', kernel_initializer='he_normal')(conv8)

    up9 = merge([UpSampling2D(size=(2, 2))(conv8), conv1], mode='concat', concat_axis=1)
    conv9 = Conv2D(32, (3, 3), activation='relu', padding='same', kernel_initializer='he_normal')(up9)

    conv10 = Conv2D(C, (1, 1), activation='relu', kernel_initializer='he_normal')(conv9)

    reshape = Reshape((C, img_h * img_w), input_shape=(C, img_h, img_w))(conv10)
    reshape = Permute((2, 1))(reshape)

    activation = Activation('softmax')(reshape)

    model = Model(input=inputs, output=activation)

    model.compile(optimizer=Adam(lr=1.0e-4), loss='categorical_crossentropy', metrics=['accuracy'])

    return model
...

后记

暂且只贴出github上的代码，待日后时间充足再详细介绍并做补充。
未完，待续…

续

  这个数据集我已经没有了，所以很多问题我没法自己实验来解答了；因为这个是当初入门时写的代码，好多地方不是很严谨，可能移植性并不是很好。
  后面会慢慢抽空更新pytorch框架下resnet，resnext等backbone的U-Net、U-Net++(Nested U-Net)、E-Net等，敬请期待…
  绝对不鸽