基于TensorFlow的手写数字识别
import struct
import numpy as np
import matplotlib.pyplot as plt
import os
from sklearn.preprocessing import OneHotEncoder
import tensorflow as tf
from PIL import Image, ImageFilter
from tensorflow.examples.tutorials.mnist import input_data
mnist = input_data.read_data_sets('MNIST_data', one_hot=True)
# 每个批次的大小
batch_size = 100
n_batch = mnist.train.num_examples
# 初始化权值
def weight_variable(shape):
inital = tf.truncated_normal(shape, stddev=0.1) # 生成一个阶段的正态分布
return tf.Variable(inital)
# 初始化偏置值
def bias_variable(shape):
inital = tf.constant(0.1, shape=shape)
return tf.Variable(inital)
# 卷积层
def conv2d(x, W):
return tf.nn.conv2d(x, W, strides=[1, 1, 1, 1], padding='SAME')
# 池化层
def max_pool_2x2(x):
return tf.nn.max_pool(x, ksize=[1, 2, 2, 1], strides=[1, 2, 2, 1], padding='SAME')
# 定义两个placeholder
# 数据类型, [有多少行,多少列]
x = tf.placeholder(tf.float32, [None, 784]) # 28*28
y = tf.placeholder(tf.float32, [None, 10])
# 改变x的格式转为4D的向量[batch, in_height, in_width, in_channels]
x_image = tf.reshape(x, [-1, 28, 28, 1])
# 初始化第一个卷积层的权值跟偏置
W_conv1 = weight_variable([5, 5, 1, 32]) # 5*5的采样窗口, 32个卷积核从1个平面抽取特征
b_conv1 = bias_variable([32]) # 每一个卷积核一个偏置值
# 把x_image 和 权值向量进行卷积,再加上偏置值,然后应用于relu激活函数
conv1 = conv2d(x_image, W_conv1) + b_conv1
h_conv1 = tf.nn.relu(conv1)
h_pool1 = max_pool_2x2(h_conv1) # 进行max_pooling
# 初始化第二个卷积层的权值跟偏置值
W_conv2 = weight_variable([5, 5, 32, 64])
b_conv2 = bias_variable([64])
# 把h_pool1 和权值向量进行卷积,再加上偏置值,然后用relu函数进行激活
conv2 = conv2d(h_pool1, W_conv2) + b_conv2
h_conv2 = tf.nn.relu(conv2)
h_pool2 = max_pool_2x2(h_conv2)
# 28 * 28 的图片第一次进行卷积后还是28*28, 第一次池化后变为14*14
# 进过第二次卷积后为14*14 第二次池化后变为7*7
# 进过上面的操作后 可以得到64张7*7的平面
# 初始化第一个全连接层的权值
W_fc1 = weight_variable([7 * 7 * 64, 1024]) # 上一场有7*7*64个神经元,全连接有1024个神经元
b_fc1 = bias_variable([1024]) # 1024个节点
# 把池化层2的输出扁平化1维
h_pool2_flat = tf.reshape(h_pool2, [-1, 7 * 7 * 64])
# 求第一个全连接层的输出
h_fc1 = tf.nn.relu(tf.matmul(h_pool2_flat, W_fc1) + b_fc1)
# keep_prob 用来表示神经元的输出概率
keep_prob = tf.placeholder(tf.float32)
h_fc1_drop = tf.nn.dropout(h_fc1, keep_prob)
# 初始化第二个全连接层
W_fc2 = weight_variable([1024, 10])
b_fc2 = bias_variable([10])
# 计算输出
prediction = tf.nn.softmax(tf.matmul(h_fc1_drop, W_fc2) + b_fc2)
# 交叉熵代价函数
cross_entropy = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(labels=y, logits=prediction))
# 使用AdamOptimizer 进行优化
train_step = tf.train.AdamOptimizer(1e-4).minimize(cross_entropy)
# 结果存放在一个布尔列表中
correct_prediction = tf.equal(tf.argmax(prediction, 1), tf.argmax(y, 1)) # argmax返回一维张量中最大值所在的位置
accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32))
"""
with tf.Session() as sess:
sess.run(tf.global_variables_initializer())
for epoch in range(21):
for batch in range(n_batch):
batch_xs, batch_ys = mnist.train.next_batch(batch_size)
testSet = mnist.test.next_batch(batch_size)
sess.run(train_step, feed_dict={x: batch_xs, y: batch_ys, keep_prob: 0.7})
acc = sess.run(accuracy, feed_dict={x: testSet[0], y: testSet[1], keep_prob: 1.0})
print("Iter" + str(epoch) + ", Testing Accuracy=" + str(acc))
"""保存模型"""
model_filepath = "./model2/gesture.ckpt"
saver = tf.train.Saver()
save_path = saver.save(sess, model_filepath)
"""
file_name = 'F:/Deep Learning/testImages/testImages/testIM6.jpg'
im = Image.open(file_name).convert('L')
img = np.array(im)
new_im = Image.fromarray(img)
new_im.show()
result = img.reshape(784, )
saver = tf.train.Saver()
with tf.Session() as sess:
sess.run(tf.global_variables_initializer())
saver.restore(sess, "./model2/gesture.ckpt")
prediction2 = tf.argmax(prediction, 1)
predint = prediction2.eval(feed_dict={x:[result], keep_prob: 1.0}, session=sess)
print('recognize result:')
print(predint)