TensorFlow study notes (eight) - TensorFlow based handwriting recognition digital data set MNIST

Training program

Referring to Tensorflow Chinese community Tutorial:
http://www.tensorfly.cn/tfdoc/tutorials/mnist_pros.html

You can successfully run the code in their own environment.
(Note that: a tutorial on the print function)

in Python3.6 version bloggers used in, print has become a function, and print them in Python2.7 is not a function, here is the need to bloggers parentheses.
Therefore, in accordance with their own version of change here, for example,

print（"Hello")  python3.6
print "Hello"   python2.7

After you follow tutorial will run through the code, the next step is to save the model, simply call a simple function saver.save, the following is based on blogger tutorial knock the complete code:

from tensorflow.examples.tutorials.mnist import input_data
import tensorflow as tf
#读取数据
mnist = input_data.read_data_sets('MNIST_data', one_hot=True)
sess=tf.InteractiveSession()
#构建cnn网络结构
#自定义卷积函数（后面卷积时就不用写太多）
def conv2d(x,w):
    return tf.nn.conv2d(x,w,strides=[1,1,1,1],padding='SAME')
#自定义池化函数
def max_pool_2x2(x):
    return tf.nn.max_pool(x,ksize=[1,2,2,1],strides=[1,2,2,1],padding='SAME')
#设置占位符，尺寸为样本输入和输出的尺寸
x=tf.placeholder(tf.float32,[None,784])
y_=tf.placeholder(tf.float32,[None,10])
x_img=tf.reshape(x,[-1,28,28,1])

#设置第一个卷积层和池化层
w_conv1=tf.Variable(tf.truncated_normal([3,3,1,32],stddev=0.1))
b_conv1=tf.Variable(tf.constant(0.1,shape=[32]))
h_conv1=tf.nn.relu(conv2d(x_img,w_conv1)+b_conv1)
h_pool1=max_pool_2x2(h_conv1)

#设置第二个卷积层和池化层
w_conv2=tf.Variable(tf.truncated_normal([3,3,32,50],stddev=0.1))
b_conv2=tf.Variable(tf.constant(0.1,shape=[50]))
h_conv2=tf.nn.relu(conv2d(h_pool1,w_conv2)+b_conv2)
h_pool2=max_pool_2x2(h_conv2)

#设置第一个全连接层
w_fc1=tf.Variable(tf.truncated_normal([7*7*50,1024],stddev=0.1))
b_fc1=tf.Variable(tf.constant(0.1,shape=[1024]))
h_pool2_flat=tf.reshape(h_pool2,[-1,7*7*50])
h_fc1=tf.nn.relu(tf.matmul(h_pool2_flat,w_fc1)+b_fc1)

#dropout（随机权重失活）
keep_prob=tf.placeholder(tf.float32)
h_fc1_drop=tf.nn.dropout(h_fc1,keep_prob)

#设置第二个全连接层
w_fc2=tf.Variable(tf.truncated_normal([1024,10],stddev=0.1))
b_fc2=tf.Variable(tf.constant(0.1,shape=[10]))
y_out=tf.nn.softmax(tf.matmul(h_fc1_drop,w_fc2)+b_fc2)

#建立loss function，为交叉熵
loss=tf.reduce_mean(-tf.reduce_sum(y_*tf.log(y_out),reduction_indices=[1]))
#配置Adam优化器，学习速率为1e-4
train_step=tf.train.AdamOptimizer(1e-4).minimize(loss)

#建立正确率计算表达式
correct_prediction=tf.equal(tf.argmax(y_out,1),tf.argmax(y_,1))
accuracy=tf.reduce_mean(tf.cast(correct_prediction,tf.float32))

# 定义saver
saver = tf.train.Saver()

#开始喂数据，训练
tf.global_variables_initializer().run()
for i in range(20000):
    batch=mnist.train.next_batch(50)
    if i%100==0:
        train_accuracy=accuracy.eval(feed_dict={x:batch[0],y_:batch[1],keep_prob:1})
        print("step %d,train_accuracy= %g"%(i,train_accuracy))
    train_step.run(feed_dict={x:batch[0],y_:batch[1],keep_prob:0.5})#这里才开始真正进行训练计算
# 模型储存位置
saver.save(sess, ".\\MNIST_data\\model.ckpt")

#训练之后，使用测试集进行测试，输出最终结果
print("test_accuracy= %g"% accuracy.eval(feed_dict={x:mnist.test.images,y_:mnist.test.labels,keep_prob:1}))

More programs for importing data set MNIST - Create Model - Save the model to the specified path. 
Saved path following four files will appear

The role of each file:

• checkpoint: text files, records the path information list of model files
• model.ckpt.data-00000-of-00001: Network Weights
• model.ckpt.index: .data and .index these two files are binary files, save the variable parameters in the model (weight) information
• model.ckpt.meta: binary file, save the configuration information of FIG calculation model (structural model of the network) Protobuf

Identification number 

Overall, just and code similarity is high, will also need to test the picture convolution, pooling, softmax return and so on. 

from PIL import Image, ImageFilter
import tensorflow as tf
import matplotlib.pyplot as plt

def imageprepare():
    im = Image.open("MNIST_data\\test3.png") #读取的图片所在路径，注意是28*28像素
    plt.imshow(im)  #显示需要识别的图片
    plt.show()
    im = im.convert('L')
    tv = list(im.getdata())
    tva = [(255-x)*1.0/255.0 for x in tv]
    return tva

x = tf.placeholder(tf.float32, [None, 784])
y_ = tf.placeholder(tf.float32, [None, 10])

def weight_variable(shape):
    initial = tf.truncated_normal(shape,stddev = 0.1)
    return tf.Variable(initial)

def bias_variable(shape):
    initial = tf.constant(0.1,shape = shape)
    return tf.Variable(initial)

def conv2d(x,W):
    return tf.nn.conv2d(x, W, strides = [1,1,1,1], padding = 'SAME')

def max_pool_2x2(x):
    return tf.nn.max_pool(x, ksize=[1,2,2,1], strides=[1,2,2,1], padding='SAME')

W_conv1 = weight_variable([3, 3, 1, 32])
b_conv1 = bias_variable([32])

x_image = tf.reshape(x,[-1,28,28,1])

h_conv1 = tf.nn.relu(conv2d(x_image,W_conv1) + b_conv1)
h_pool1 = max_pool_2x2(h_conv1)

W_conv2 = weight_variable([3, 3, 32, 50])
b_conv2 = bias_variable([50])

h_conv2 = tf.nn.relu(conv2d(h_pool1, W_conv2) + b_conv2)
h_pool2 = max_pool_2x2(h_conv2)

W_fc1 = weight_variable([7 * 7 * 50, 1024])
b_fc1 = bias_variable([1024])

h_pool2_flat = tf.reshape(h_pool2, [-1, 7*7*50])
h_fc1 = tf.nn.relu(tf.matmul(h_pool2_flat, W_fc1) + b_fc1)

keep_prob = tf.placeholder("float")
h_fc1_drop = tf.nn.dropout(h_fc1, keep_prob)

W_fc2 = weight_variable([1024, 10])
b_fc2 = bias_variable([10])

y_conv=tf.nn.softmax(tf.matmul(h_fc1_drop, W_fc2) + b_fc2)

cross_entropy = -tf.reduce_sum(y_*tf.log(y_conv))
train_step = tf.train.AdamOptimizer(1e-4).minimize(cross_entropy)
correct_prediction = tf.equal(tf.argmax(y_conv,1), tf.argmax(y_,1))
accuracy = tf.reduce_mean(tf.cast(correct_prediction, "float"))

saver = tf.train.Saver()

with tf.Session() as sess:
    sess.run(tf.global_variables_initializer())
    saver.restore(sess, "MNIST_data\\model.ckpt") #使用模型，参数和之前的代码保持一致

    result = imageprepare()

    prediction=tf.argmax(y_conv,1)
    predint=prediction.eval(feed_dict={x: [result],keep_prob: 1.0}, session=sess)

    print('识别结果:')
    print(predint[0])

From the test results, the failure rate is still pricey.

summary of a problem 

Create an image

 Image size must be the X-28 28 , or you can choose ps beauties Xiu Xiu or any other drawing tools.

The following are the bloggers made by Mito Xiu Xiu Pictures

 error:

tensorflow.python.framework.errors_impl.InvalidArgumentError: Assign requires shapes of both tensors to match

 This error is due to do convolution and pooled in the training model when selecting the size of inconsistencies caused.

for example:

Save the model used is

W_conv2 = weight_variable([3, 3, 32, 50])

But when use is identified

W_conv2 = weight_variable([5, 5, 32, 64])

Be sure to keep the same parameters