语义分割

1.vgg_segnet

from keras.models import *
from keras.layers import *
from keras.activations import *
import keras.backend as K
import keras
IMAGE_ORDERRING='channels_last'
#1.encoder_0.5net
def convnet_encoder(input_height=416,input_width=416,pretained='imagenet'):
    img_input=Input(shape=(input_height,input_width,3))
    x=Conv2D(64,(3,3),activation='relu',padding='same',name='block1_conv1')(img_input)
    x=Conv2D(64,(3,3),activation='relu',padding='same',name='block1_conv2')(x)
    x=MaxPooling2D((2,2),strides=(2,2),name='block_pool')(x)
    f1=x
    x=Conv2D(128,(3,3),activation='relu',padding='same',name='block2_conv1')(x)
    x=Conv2D(128,(3,3),activation='relu',padding='same',name='block2_conv2')(x)
    x=MaxPooling2D((2,2),strides=(2,2),name='block2_pool')(x)
    f2=x
    x=Conv2D(256,(3,3),activation='relu',padding='same',name='block3_conv1')(x)
    x=Conv2D(256,(3,3),activation='relu',padding='same',name='block3_conv2')(x)
    x=Conv2D(256,(3,3),activation='relu',padding='same',name='block3_conv3')(x)
    x=MaxPooling2D((2,2),strides=(2,2),name='block3_pool')(x)
    f3=x
    x=Conv2D(512,(3,3),activation='relu',padding='same',name='block4_conv1')(x)
    x=Conv2D(512,(3,3),activation='relu',padding='same',name='block4_conv2')(x)
    x=Conv2D(512,(3,3),activation='relu',padding='same',name='block4_conv3')(x)
    x=MaxPooling2D((2,2),strides=(2,2),name='block4_pool')(x)
    f4=x
    x=Conv2D(1024,(3,3),activation='relu',padding='same',name='block5_conv1')(x)
    x=Conv2D(1024,(3,3),activation='relu',padding='same',name='block5_conv2')(x)
    x=Conv2D(1024,(3,3),activation='relu',padding='same',name='block5_conv3')(x)
    x=MaxPooling2D((2,2),strides=(2,2),name='block5_pool')(x)
    f5=x
    return img_input,[f1,f2,f3,f4,f5]

　　

#2.decoder_0.5net
def segnet_decoder(output_input,n_classes):
    x=ZeroPadding2D((1,1))(output_input)
    x=Conv2D(512,(3,3),padding='valid')(x)
    x=BatchNormalization()(x)
    x=UpSampling2D((2,2))(x)
    x=ZeroPadding2D((1,1))(x)
    x=Conv2D(256,(3,3),padding='valid')(x)
    x=BatchNormalization()(x)
    x=UpSampling2D((2,2))(x)
    x=ZeroPadding2D((1,1))(x)
    x=Conv2D(128,(3,3),padding='valid')(x)
    x=BatchNormalization()(x)
    x=UpSampling2D((2,2) )(x)
    x=ZeroPadding2D((1,1))(x)
    x=Conv2D(64,(3,3),padding='valid')(x)
    x=BatchNormalization()(x)
    x=Conv2D(n_classes,(3,3),padding='same')(x)
    return x

　　

#3.vgg_segnet_net
def convnet_segnet(n_classes,input_height=416,input_width=416):
    img_input,levels=convnet_encoder(input_height=input_height,input_width=input_width)
    feat=levels[3]#f4
    print(feat.shape)
    x=segnet_decoder(feat,n_classes)
    #将结果reshape到2维，就是每个像素点的预测类别
    x=Reshape((int(input_height/2)*int(input_width/2),-1))(x)#(每个像素点，类别onehot)
    out=Softmax()(x)#概率最大类别
    model=Model(img_input,out)#构建模型（输入输出）
    model.model_name='convnet_segnet'
    return model
#查看网络结构
model=convnet_segnet(2,input_height=416,input_width=416)
model.summary()