python识别验证码tensorflow_Python Tensorflow CNN 识别验证码

Python+Tensorflow的CNN技术快速识别验证码

文章来源于： https://www.jianshu.com/p/26ff7b9075a1

验证码处理的流程是：验证码分析和处理—— tensorflow安装 —— 模型训练 —— 模型预测

需要的准备。

1. 安装TensorFlow

2.  PIL

3. numpy

4. 用于训练的图片

0.文件目录：

红色部分有用，其他不用

1. 训练模型的图片：链接：https://pan.baidu.com/s/1kpgt7Pc-ni4WnN6qj8U-pw 密码：nzea

2.  训练模型代码：

训练好的模型：链接：https://pan.baidu.com/s/1dNpEtguITKBgbsUU6tCluQ 密码：j07f

from PIL importImageimportnumpy as npimporttensorflow as tfimportos

os.environ['TF_CPP_MIN_LOG_LEVEL']='2'

importrandom

IMAGE_HEIGHT= 114IMAGE_WIDTH= 450MAX_CAPTCHA= 6CHAR_SET_LEN= 26

defget_name_and_image():

all_image= os.listdir('C:\\Users\\xuchunlin\\PycharmProjects\\ML\\20180402\\captcha4\\')

random_file= random.randint(0, 3429)

base= os.path.basename('C:\\Users\\xuchunlin\\PycharmProjects\\ML\\20180402\\captcha4\\' +all_image[random_file])

name=os.path.splitext(base)[0]

image= Image.open('C:\\Users\\xuchunlin\\PycharmProjects\\ML\\20180402\\captcha4\\' +all_image[random_file])

image=np.array(image)returnname, imagedefname2vec(name):

vector= np.zeros(MAX_CAPTCHA*CHAR_SET_LEN)for i, c inenumerate(name):

idx= i * 26 + ord(c) - 97vector[idx]= 1

returnvectordefvec2name(vec):

name=[]for i invec:

a= chr(i + 97)

name.append(a)return "".join(name)#生成一个训练batch

def get_next_batch(batch_size=64):

batch_x= np.zeros([batch_size, IMAGE_HEIGHT*IMAGE_WIDTH])

batch_y= np.zeros([batch_size, MAX_CAPTCHA*CHAR_SET_LEN])for i inrange(batch_size):

name, image=get_name_and_image()

batch_x[i, :]= 1*(image.flatten())

batch_y[i, :]=name2vec(name)returnbatch_x, batch_y####################################################

X= tf.placeholder(tf.float32, [None, IMAGE_HEIGHT*IMAGE_WIDTH])

Y= tf.placeholder(tf.float32, [None, MAX_CAPTCHA*CHAR_SET_LEN])

keep_prob=tf.placeholder(tf.float32)#定义CNN

def crack_captcha_cnn(w_alpha=0.01, b_alpha=0.1):

x= tf.reshape(X, shape=[-1, IMAGE_HEIGHT, IMAGE_WIDTH, 1])#3 conv layer

w_c1 = tf.Variable(w_alpha * tf.random_normal([5, 5, 1, 32]))

b_c1= tf.Variable(b_alpha * tf.random_normal([32]))

conv1= tf.nn.relu(tf.nn.bias_add(tf.nn.conv2d(x, w_c1, strides=[1, 1, 1, 1], padding='SAME'), b_c1))

conv1= tf.nn.max_pool(conv1, ksize=[1, 2, 2, 1], strides=[1, 2, 2, 1], padding='SAME')

conv1=tf.nn.dropout(conv1, keep_prob)

w_c2= tf.Variable(w_alpha * tf.random_normal([5, 5, 32, 64]))

b_c2= tf.Variable(b_alpha * tf.random_normal([64]))

conv2= tf.nn.relu(tf.nn.bias_add(tf.nn.conv2d(conv1, w_c2, strides=[1, 1, 1, 1], padding='SAME'), b_c2))

conv2= tf.nn.max_pool(conv2, ksize=[1, 2, 2, 1], strides=[1, 2, 2, 1], padding='SAME')

conv2=tf.nn.dropout(conv2, keep_prob)

w_c3= tf.Variable(w_alpha * tf.random_normal([5, 5, 64, 64]))

b_c3= tf.Variable(b_alpha * tf.random_normal([64]))

conv3= tf.nn.relu(tf.nn.bias_add(tf.nn.conv2d(conv2, w_c3, strides=[1, 1, 1, 1], padding='SAME'), b_c3))

conv3= tf.nn.max_pool(conv3, ksize=[1, 2, 2, 1], strides=[1, 2, 2, 1], padding='SAME')

conv3=tf.nn.dropout(conv3, keep_prob)#Fully connected layer

w_d = tf.Variable(w_alpha * tf.random_normal([15 * 57 * 64, 1024]))

b_d= tf.Variable(b_alpha * tf.random_normal([1024]))

dense= tf.reshape(conv3, [-1, w_d.get_shape().as_list()[0]])

dense=tf.nn.relu(tf.add(tf.matmul(dense, w_d), b_d))

dense=tf.nn.dropout(dense, keep_prob)

w_out= tf.Variable(w_alpha * tf.random_normal([1024, MAX_CAPTCHA *CHAR_SET_LEN]))

b_out= tf.Variable(b_alpha * tf.random_normal([MAX_CAPTCHA *CHAR_SET_LEN]))

out=tf.add(tf.matmul(dense, w_out), b_out)returnout#训练

deftrain_crack_captcha_cnn():

output=crack_captcha_cnn()

loss= tf.reduce_mean(tf.nn.sigmoid_cross_entropy_with_logits(logits=output, labels=Y))

optimizer= tf.train.AdamOptimizer(learning_rate=0.001).minimize(loss)

predict= tf.reshape(output, [-1, MAX_CAPTCHA, CHAR_SET_LEN])

max_idx_p= tf.argmax(predict, 2)

max_idx_l= tf.argmax(tf.reshape(Y, [-1, MAX_CAPTCHA, CHAR_SET_LEN]), 2)

correct_pred=tf.equal(max_idx_p, max_idx_l)

accuracy=tf.reduce_mean(tf.cast(correct_pred, tf.float32))

saver=tf.train.Saver()

with tf.Session() as sess:

sess.run(tf.global_variables_initializer())

step=0whileTrue:

batch_x, batch_y= get_next_batch(64)

_, loss_= sess.run([optimizer, loss], feed_dict={X: batch_x, Y: batch_y, keep_prob: 0.5})print(step, loss_)#每100 step计算一次准确率

if step % 100 ==0:

batch_x_test, batch_y_test= get_next_batch(100)

acc= sess.run(accuracy, feed_dict={X: batch_x_test, Y: batch_y_test, keep_prob: 1.})print(step, acc)#如果准确率大于60%,保存模型,完成训练

if acc > 0.6:

saver.save(sess,"./crack_capcha.model", global_step=step)breakstep+= 1train_crack_captcha_cnn()

3.  模型测试代码：

defcrack_captcha():

output=crack_captcha_cnn()

saver=tf.train.Saver()

with tf.Session() as sess:

saver.restore(sess, tf.train.latest_checkpoint('.'))

n= 1

while n <= 10:

text, image=get_name_and_image()

image= 1 *(image.flatten())

predict= tf.argmax(tf.reshape(output, [-1, MAX_CAPTCHA, CHAR_SET_LEN]), 2)

text_list= sess.run(predict, feed_dict={X: [image], keep_prob: 1})

vec=text_list[0].tolist()

predict_text=vec2name(vec)print("正确: {} 预测: {}".format(text, predict_text))

n+= 1crack_captcha()

训练代码和测试代码文件： 链接：https://pan.baidu.com/s/1VY9rYZizCEjHzim3-XaGyw 密码：epv2

结果展示：

你会发现识别率并不高，那是因为上面训练模型中有这几行代码

#如果准确率大于60%,保存模型,完成训练

if acc > 0.6:

saver.save(sess,"./crack_capcha.model", global_step=step)break

设定的准确率只有百分之六十，如果时间充足的话，可以设置0.99或者0.98.会得到一个不错的模型。

详细讲解请去原网址看，地址：https://www.jianshu.com/p/26ff7b9075a1

所有学习资料：链接：https://pan.baidu.com/s/19BoO5sUhLrzpL0a9_rNTRQ 密码：q4ri