帮我写一个unet网络，我输入的图像为3*384*384，希望输出也是3*384*384

UNet是一种经典的卷积神经网络架构，常用于医学图像分割、遥感图像分析等领域。它特别适合处理像你所描述的大尺寸图像，并通过上下采样操作将低分辨率特征图与高分辨率输入对齐。以下是基本的Python代码片段，使用Keras库来构建一个适用于3通道384x384图像的简化的UNet模型：

import tensorflow as tf
from tensorflow.keras.layers import Input, Conv2D, MaxPooling2D, UpSampling2D, Concatenate

def unet(input_size=(3, 384, 384)):
    inputs = Input(input_size)

    # 编码部分 (Downsampling)
    conv1 = Conv2D(64, (3, 3), activation='relu', padding='same')(inputs)
    pool1 = MaxPooling2D(pool_size=(2, 2))(conv1)
    conv2 = Conv2D(128, (3, 3), activation='relu', padding='same')(pool1)
    pool2 = MaxPooling2D(pool_size=(2, 2))(conv2)
    conv3 = Conv2D(256, (3, 3), activation='relu', padding='same')(pool2)
    pool3 = MaxPooling2D(pool_size=(2, 2))(conv3)

    # 中间层
    conv4 = Conv2D(512, (3, 3), activation='relu', padding='same')(pool3)

    # 解码部分 (Upsampling)
    up5 = Conv2DTranspose(256, (2, 2), strides=(2, 2), padding='same')(conv4)  # Deconvolution or UpSampling + concatenation
    merge5 = Concatenate(axis=3)([up5, conv3])
    conv5 = Conv2D(256, (3, 3), activation='relu', padding='same')(merge5)

    up6 = Conv2DTranspose(128, (2, 2), strides=(2, 2), padding='same')(conv5)
    merge6 = Concatenate(axis=3)([up6, conv2])
    conv6 = Conv2D(128, (3, 3), activation='relu', padding='same')(merge6)

    up7 = Conv2DTranspose(64, (2, 2), strides=(2, 2), padding='same')(conv6)
    merge7 = Concatenate(axis=3)([up7, conv1])
    conv7 = Conv2D(64, (3, 3), activation='relu', padding='same')(merge7)

    # 输出层
    output = Conv2D(3, (1, 1), activation='sigmoid')(conv7)  # 这里假设输出需要3通道概率图

    model = tf.keras.Model(inputs=[inputs], outputs=[output])

    return model

# 使用model.summary()检查模型结构

请注意，这只是一个基础版本的UNet，实际应用中你可能需要调整层数、滤波器数量等超参数以适应特定任务。同时，如果你的数据集较大，可能还需要添加数据增强和批归一化步骤。