Unet网络架构详解

创作时间:

作者:

@小白创作中心

Unet网络架构详解

引用

CSDN

https://blog.csdn.net/weixin_40641178/article/details/143460158

Unet是一种用于图像分割的深度学习网络架构，其核心特点是通过编码器-解码器结构实现特征的提取和还原，并通过"copy and crop"操作实现高分辨率特征图和低分辨率特征图的融合。本文将从零开始，详细讲解Unet网络的实现过程。

整体架构图

网络讲解

编码器部分

Unet的编码器部分由多个卷积块组成，每个卷积块包含两个卷积层和一个ReLU激活函数，最后接一个最大池化层。具体实现如下：

class DoubleConv(nn.Module):
    """(convolution => [BN] => ReLU) * 2"""
    def __init__(self, in_channels, out_channels, mid_channels=None):
        super().__init__()
        if not mid_channels:
            mid_channels = out_channels
        self.double_conv = nn.Sequential(
            nn.Conv2d(in_channels, mid_channels, kernel_size=3, padding=1),
            nn.BatchNorm2d(mid_channels),
            nn.ReLU(inplace=True),
            nn.Conv2d(mid_channels, out_channels, kernel_size=3, padding=1),
            nn.BatchNorm2d(out_channels),
            nn.ReLU(inplace=True)
        )
    def forward(self, x):
        return self.double_conv(x)

class Down(nn.Module):
    """Downscaling with maxpool then double conv"""
    def __init__(self, in_channels, out_channels):
        super().__init__()
        self.maxpool_conv = nn.Sequential(
            nn.MaxPool2d(2),
            DoubleConv(in_channels, out_channels)
        )
    def forward(self, x):
        return self.maxpool_conv(x)

解码器部分

解码器部分主要由上采样层和卷积层组成。上采样层可以使用转置卷积或双线性插值，然后通过卷积层进行特征融合。具体实现如下：

class Up(nn.Module):
    """Upscaling then double conv"""
    def __init__(self, in_channels, out_channels, bilinear=True):
        super().__init__()
        # if bilinear, use the normal convolutions to reduce the number of channels
        if bilinear:
            self.up = nn.Upsample(scale_factor=2, mode='bilinear', align_corners=True)  # 双线性插值
            self.conv = DoubleConv(in_channels, out_channels, in_channels // 2)
        else:
            self.up = nn.ConvTranspose2d(in_channels, in_channels // 2, kernel_size=2, stride=2)  # 转置卷积
            self.conv = DoubleConv(in_channels, out_channels)
    def forward(self, x1, x2):
        x1 = self.up(x1)
        # input is CHW
        diffY = x2.size()[2] - x1.size()[2]
        diffX = x2.size()[3] - x1.size()[3]
        x1 = F.pad(x1, [diffX // 2, diffX - diffX // 2,
                        diffY // 2, diffY - diffY // 2])
        x = torch.cat([x2, x1], dim=1)
        return self.conv(x)

输出层

最后通过一个1x1的卷积层输出最终的分割结果：

class OutConv(nn.Module):
    def __init__(self, in_channels, out_channels):
        super(OutConv, self).__init__()
        self.conv = nn.Conv2d(in_channels, out_channels, kernel_size=1)
    def forward(self, x):
        return self.conv(x)

整体网络结构

将上述组件组合成完整的Unet网络：

class UNet(nn.Module):
    def __init__(self, args, n_channels, n_classes, bilinear=True):
        super(UNet, self).__init__()  # 简单点讲：就是子类使用父类的初始化方法进行初始化，这会使得代码非常的整洁
        self.n_channels = n_channels
        self.n_classes = n_classes
        self.bilinear = bilinear
        """DoubleConv <-> (convolution => [BN] => ReLU) * 2"""
        self.inc = DoubleConv(n_channels, 64)
        self.down1 = Down(64, 128)
        self.down2 = Down(128, 256)
        self.down3 = Down(256, 512)
        factor = 2 if bilinear else 1
        self.down4 = Down(512, 1024 // factor)
        self.up1 = Up(1024, 512 // factor, bilinear)
        self.up2 = Up(512, 256 // factor, bilinear)
        self.up3 = Up(256, 128 // factor, bilinear)
        self.up4 = Up(128, 64, bilinear)
        self.outc = OutConv(64, n_classes)
    def forward(self, x):
        x1 = self.inc(x)
        x2 = self.down1(x1)
        x3 = self.down2(x2)
        x4 = self.down3(x3)
        x5 = self.down4(x4)
        x = self.up1(x5, x4)
        x = self.up2(x, x3)
        x = self.up3(x, x2)
        x = self.up4(x, x1)
        logits = self.outc(x)
        return logits

"Copy and Crop"操作

在Unet的解码器部分，需要将上采样得到的特征图与编码器对应位置的特征图进行拼接。由于上采样后的特征图尺寸可能与编码器特征图尺寸不匹配，因此需要进行"copy and crop"操作，即对编码器特征图进行中心裁剪，使其与上采样特征图尺寸一致，然后再进行拼接。

# 中心裁剪函数
def crop_tensor(self, tensor, target_tensor):
    target_size = target_tensor.size()[2]
    tensor_size = tensor.size()[2]
    delta = tensor_size - target_size
    delta = delta // 2
    return tensor[:, :, delta:tensor_size - delta, delta:tensor_size - delta]

# 拼接操作示例
up1 = self.up_conv_1(x10)  # 得到56*56*512
crop1 = self.crop_tensor(x8, up1)  # 对x8进行裁剪
up_1 = torch.cat([crop1, up1], dim=1)  # 拼接操作