我们如何通过 AI 模拟实现属于自己的清明上河图?

作者 | 李秋键

责编 | 刘静

出品 | CSDN(ID:CSDNnews)

我们知道清明上河图是我国国画的代表作之一,是中国十大传世名画 之一。为北宋风俗画 ,北宋画家张择端 仅见的存世精品,属国宝级文物 ,现藏于北京故宫博物院。

清明上河图宽24.8厘米、长528.7厘米 ,绢本设色 。作品以长卷 形式,采用散点透视 构图法,生动记录了中国十二世纪北宋 都城东京(又称汴京 ,今河南开封 )的城市面貌和当时社会各阶层人民的生活状况,是北宋时期都城汴京当年繁荣的见证,也是北宋城市经济情况的写照。

这在中国乃至世界绘画史上都是独一无二的。在五米多长的画卷里,共绘了数量庞大的各色人物,牛、骡、驴等牲畜,车、轿、大小船只,房屋、桥梁、城楼 等各有特色,体现了宋代建筑的特征。具有很高的历史价值和艺术价值。《清明上河图》虽然场面热闹,但表现的并非繁荣市景,而是一幅带有忧患意识的"盛世危图",官兵懒散税务重。

而我们今天的项目就是通过对算法的改造,实现属于自己的清明上河图。

下面我们将利用vgg19模型训练画作,详细步骤如下,并且我在每个代码上面都注释了方便查看:

首先我们导入先关的库:

import tensorflow as tf

import numpy as np

import scipy.io

import scipy.misc

import os

import time

接着定义一些变量方便调用:

CONTENT_IMG = '1.png'

STYLE_IMG = 'sty.jpg'

OUTPUT_DIR = 'neural_style_transfer_tensorflow/'

再创建一个目录用来保存图片:

if not os.path.exists(OUTPUT_DIR):

os.mkdir(OUTPUT_DIR)

定义生成图像的长宽通道等信息:

IMAGE_W = 400

IMAGE_H = 300

COLOR_C = 3

NOISE_RATIO = 0.7

BETA = 5

ALPHA = 100

再接着定义模型路径

VGG_MODEL = 'imagenet-vgg-verydeep-19.mat'

生成一个参数矩阵,作为图像的处理过程之一,对像素值运算:

MEAN_VALUES = np.array([123.68, 116.779, 103.939]).reshape((1, 1, 1, 3))

再接着定义读取模型函数,下面我都有所注解:

def load_vgg_model(path):

'''

Details of the VGG19 model:

- 0 is conv1_1 (3, 3, 3, 64)

- 1 is relu

- 2 is conv1_2 (3, 3, 64, 64)

- 3 is relu

- 4 is maxpool

- 5 is conv2_1 (3, 3, 64, 128)

- 6 is relu

- 7 is conv2_2 (3, 3, 128, 128)

- 8 is relu

- 9 is maxpool

- 10 is conv3_1 (3, 3, 128, 256)

- 11 is relu

- 12 is conv3_2 (3, 3, 256, 256)

- 13 is relu

- 14 is conv3_3 (3, 3, 256, 256)

- 15 is relu

- 16 is conv3_4 (3, 3, 256, 256)

- 17 is relu

- 18 is maxpool

- 19 is conv4_1 (3, 3, 256, 512)

- 20 is relu

- 21 is conv4_2 (3, 3, 512, 512)

- 22 is relu

- 23 is conv4_3 (3, 3, 512, 512)

- 24 is relu

- 25 is conv4_4 (3, 3, 512, 512)

- 26 is relu

- 27 is maxpool

- 28 is conv5_1 (3, 3, 512, 512)

- 29 is relu

- 30 is conv5_2 (3, 3, 512, 512)

- 31 is relu

- 32 is conv5_3 (3, 3, 512, 512)

- 33 is relu

- 34 is conv5_4 (3, 3, 512, 512)

- 35 is relu

- 36 is maxpool

- 37 is fullyconnected (7, 7, 512, 4096)

- 38 is relu

- 39 is fullyconnected (1, 1, 4096, 4096)

- 40 is relu

- 41 is fullyconnected (1, 1, 4096, 1000)

- 42 is softmax

'''

vgg = scipy.io.loadmat(path)

vgg_layers = vgg['layers']

#加载vgg模型获取模型各层参数和名称

def _weights(layer, expected_layer_name):

W = vgg_layers[0][layer][0][0][2][0][0]

b = vgg_layers[0][layer][0][0][2][0][1]

layer_name = vgg_layers[0][layer][0][0][0][0]

assert layer_name == expected_layer_name

return W, b

#将加载的变量初始化成tf可运算的张量类型,函数返回值为激活函数的输出

def _conv2d_relu(prev_layer, layer, layer_name):

W, b = _weights(layer, layer_name)

W = tf.constant(W)

b = tf.constant(np.reshape(b, (b.size)))

return tf.nn.relu(tf.nn.conv2d(prev_layer, filter=W, strides=[1, 1, 1, 1], padding='SAME') + b)

#定义池化层函数

def _avgpool(prev_layer):

return tf.nn.avg_pool(prev_layer, ksize=[1, 2, 2, 1], strides=[1, 2, 2, 1], padding='SAME')

#将各层输出值都放到列表中方便加载,形成字典

graph = {}

graph['input'] = tf.Variable(np.zeros((1, IMAGE_H, IMAGE_W, COLOR_C)), dtype='float32')

#定义['conv1_1']为vgg模型的第0层,输入层为上一层的['input' ]

graph['conv1_1'] = _conv2d_relu(graph['input'], 0, 'conv1_1')

graph['conv1_2'] = _conv2d_relu(graph['conv1_1'], 2, 'conv1_2')

graph['avgpool1'] = _avgpool(graph['conv1_2'])

graph['conv2_1'] = _conv2d_relu(graph['avgpool1'], 5, 'conv2_1')

graph['conv2_2'] = _conv2d_relu(graph['conv2_1'], 7, 'conv2_2')

graph['avgpool2'] = _avgpool(graph['conv2_2'])

graph['conv3_1'] = _conv2d_relu(graph['avgpool2'], 10, 'conv3_1')

graph['conv3_2'] = _conv2d_relu(graph['conv3_1'], 12, 'conv3_2')

graph['conv3_3'] = _conv2d_relu(graph['conv3_2'], 14, 'conv3_3')

graph['conv3_4'] = _conv2d_relu(graph['conv3_3'], 16, 'conv3_4')

graph['avgpool3'] = _avgpool(graph['conv3_4'])

graph['conv4_1'] = _conv2d_relu(graph['avgpool3'], 19, 'conv4_1')

graph['conv4_2'] = _conv2d_relu(graph['conv4_1'], 21, 'conv4_2')

graph['conv4_3'] = _conv2d_relu(graph['conv4_2'], 23, 'conv4_3')

graph['conv4_4'] = _conv2d_relu(graph['conv4_3'], 25, 'conv4_4')

graph['avgpool4'] = _avgpool(graph['conv4_4'])

graph['conv5_1'] = _conv2d_relu(graph['avgpool4'], 28, 'conv5_1')

graph['conv5_2'] = _conv2d_relu(graph['conv5_1'], 30, 'conv5_2')

graph['conv5_3'] = _conv2d_relu(graph['conv5_2'], 32, 'conv5_3')

graph['conv5_4'] = _conv2d_relu(graph['conv5_3'], 34, 'conv5_4')

graph['avgpool5'] = _avgpool(graph['conv5_4'])

return graph

为了实现自己的项目效果,设定损失函数:

#定义内容损失函数,变量为tf计算图和vgg模型参数,返回值为损失值

def content_loss_func(sess, model):

#p就是model['conv4_2'])参数,x是model['conv4_2'])

def _content_loss(p, x):

#p的值为Tensor("Relu_9:0", shape=(1, 75, 100, 512), dtype=float32),故N为512,M为75*100,分别为卷积核个数,卷积核大小的宽*100

N = p.shape[3]

M = p.shape[1] * p.shape[2]

return (1 / (4 * N * M)) * tf.reduce_sum(tf.pow(x - p, 2))

return _content_loss(sess.run(model['conv4_2']), model['conv4_2'])

STYLE_LAYERS = [('conv1_1', 0.5), ('conv2_1', 1.0), ('conv3_1', 1.5), ('conv4_1', 3.0), ('conv5_1', 4.0)]

#返回值为_style_loss的值*0.5,1,1.5,4的加和

def style_loss_func(sess, model):

def _gram_matrix(F, N, M):

Ft = tf.reshape(F, (M, N))

return tf.matmul(tf.transpose(Ft), Ft)

#a,x都为'conv1_1', conv2_1', 'conv3_1', 'conv4_1','conv5_1'中的参数遍历

def _style_loss(a, x):

#同内容损失函数

N = a.shape[3]

M = a.shape[1] * a.shape[2]

A = _gram_matrix(a, N, M)

G = _gram_matrix(x, N, M)

return (1 / (4 * N ** 2 * M ** 2)) * tf.reduce_sum(tf.pow(G - A, 2))

return sum([_style_loss(sess.run(model[layer_name]), model[layer_name]) * w for layer_name, w in STYLE_LAYERS])

再定义生成图片,读取图片,保存图片函数:

#产生噪声图片

def generate_noise_image(content_image, noise_ratio=NOISE_RATIO):

#随机产生矩阵图片,矩阵元素内容符合标准正太分布

noise_image = np.random.uniform(-20, 20, (1, IMAGE_H, IMAGE_W, COLOR_C)).astype('float32')

#将产生的矩阵内各元素与神经网络加和

input_image = noise_image * noise_ratio + content_image * (1 - noise_ratio)

return input_image

#读取图片,改变尺寸,变成1行多列矩阵,将矩阵与初始值相减返回

def load_image(path):

image = scipy.misc.imread(path)

image = scipy.misc.imresize(image, (IMAGE_H, IMAGE_W))

#image.shape为[800,600,3],则(1, ) + image.shape)为[1,800,600,3]

image = np.reshape(image, ((1, ) + image.shape))

#MEAN_VALUES = np.array([123.68, 116.779, 103.939]).reshape((1, 1, 1, 3))

#其中image为三通道矩阵,MEAN_VALUES为三维矩阵可以相减

image = image - MEAN_VALUES

return image

#保存图片

def save_image(path, image):

image = image + MEAN_VALUES

#参见上面图像加载时多加了1维,故形成时要减少维度,

image = image[0]

#截取所有数值在0-255之间的,因为像素值必须是这个范围。而参数运算后可能会超过这个值

image = np.clip(image, 0, 255).astype('uint8')

#保存

scipy.misc.imsave(path, image)

下面是训练加载:

#启动计算图

with tf.Session() as sess:

#读取图片,返回值为减去MEAN_VALUES的矩阵,矩阵形状为[1,800,600,3]

content_image = load_image(CONTENT_IMG)

style_image = load_image(STYLE_IMG)

#加载vgg19模型,返回值为一个字典,里面为各网络层参数,输入和输出

model = load_vgg_model(VGG_MODEL)

#产生噪声图片,返回值为随机矩阵加上网络层参数的新矩阵

input_image = generate_noise_image(content_image)

#变量初始化

sess.run(tf.global_variables_initializer())

#从网络层input层开始运算内容图片矩阵

sess.run(model['input'].assign(content_image))

content_loss = content_loss_func(sess, model)

# 从网络层input层开始运算内容图片矩阵

sess.run(model['input'].assign(style_image))

style_loss = style_loss_func(sess, model)

#总损失为内容损失加上风格损失

total_loss = BETA * content_loss + ALPHA * style_loss

#建立优化器以调整参数

optimizer = tf.train.AdamOptimizer(2.0)

#优化器调整参数,使得损失为最小

train = optimizer.minimize(total_loss)

sess.run(tf.global_variables_initializer())

# 从网络层input层开始运算形成新的图片

sess.run(model['input'].assign(input_image))

ITERATIONS = 2000

#训练2000轮

for i in range(ITERATIONS):

sess.run(train)

print('Iteration %d' % i)

print('Cost: ', sess.run(total_loss))

if i % 100 == 0:

#每一百次加载一次网络参数以保存图片

output_image = sess.run(model['input'])

print('Iteration %d' % i)

print('Cost: ', sess.run(total_loss))

save_image(os.path.join(OUTPUT_DIR, 'output_%d.jpg' % i), output_image)

最终得到的效果如图所示:

左边是电视里找的图片,右边是模拟的图片,由此可见生成的效果还是可以的。而这个程序的主要思路就是在一个生成随机矩阵的基础上,通过加载网络层训练参数,然后生成的矩阵值按比例乘以网络参数,然后把矩阵保存为图片即可达到模拟生成的效果。而其中参数的调整是基于深层次网络提取的图像特征按公式运算,通过优化器优化参数,通过训练次数的增加,参数也在逐渐改善,最终形成自己需要的图片效果。

作者简介:李秋键,CSDN 博客专家,CSDN达人课作者。硕士在读于中国矿业大学,开发有安卓武侠游戏一部,VIP视频解析,文意转换写作机器人等项目,发表论文若干,多次高数竞赛获奖等等。

声明:本文为作者原创投稿,未经允许请勿转载。

【END】