TensorFlow_CNN_MNIST

# -*- coding: utf-8 -*- """ Created on Sat Nov 25 22:08:35 2017 @author: Administrator """ import tensorflow as tf from tensorflow.examples.tutorials.mnist import input_data mnist = input_data.read_data_sets('MNIST_data', one_hot=True) import tensorflow as tf import numpy import os from PIL import Image from skimage import img_as_int from skimage import img_as_ubyte import cv2 ''' 导入图像 ''' im = Image.open("C:/Users/Administrator/Desktop/ml_test/my_test_data/my_test0_6.jpg") #im = Image.open("3.jpg") #转换灰度图 im_L =numpy.mat(im.convert("L")) im_dL=255-img_as_ubyte(im_L) im_dL=numpy.float32(im_dL/255) data_to_test= numpy.mat(im_dL[0,:]) for i in range(1,28): data_to_test=numpy.hstack ((data_to_test,numpy.mat(im_dL[i,:]))) #the_beging_data=tf.convert_to_tensor(data_to_test) #data_to_test=numpy.mat(mnist.test.images[0,:]) my_y_=numpy.array([[1,0,0,0,0,0,0,0,0,0]]) sess = tf.InteractiveSession() x = tf.placeholder("float", shape=[None, 784]) y_ = tf.placeholder("float", shape=[None, 10]) W = tf.Variable(tf.zeros([784,10])) b = tf.Variable(tf.zeros([10])) #sess.run(tf.initialize_all_variables()) #y = tf.nn.softmax(tf.matmul(x,W) + b) ''' 权重初始化 ''' 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) ''' 卷积和池化 ''' 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') #第一层卷积 W_conv1 = weight_variable([5, 5, 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) #第二层卷积 W_conv2 = weight_variable([5, 5, 32, 64]) b_conv2 = bias_variable([64]) h_conv2 = tf.nn.relu(conv2d(h_pool1, W_conv2) + b_conv2) h_pool2 = max_pool_2x2(h_conv2) #连接层 W_fc1 = weight_variable([7 * 7 * 64, 1024]) b_fc1 = bias_variable([1024]) h_pool2_flat = tf.reshape(h_pool2, [-1, 7*7*64]) 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]) ''' 构建模型，交叉熵 ''' 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")) ''' 初始化开始训练 ''' sess.run(tf.initialize_all_variables()) """ 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: 1.0}) print ("step %d, training accuracy %g"%(i, train_accuracy)) train_step.run(feed_dict={x: batch[0], y_: batch[1], keep_prob: 0.5}) """ saver = tf.train.Saver({'W_conv1':W_conv1,'b_conv1':b_conv1,'W_conv2':W_conv2,'b_conv2':b_conv2,'W_fc1':W_fc1,'b_fc1':b_fc1,'W_fc2':W_fc2,'b_fc2':b_fc2}) sess = tf.InteractiveSession() saver.restore(sess,'C:/Users/Administrator/Desktop/ml_test/Train_result/Result') ''' 打印结果 '''''' print ("test accuracy %g"%accuracy.eval(feed_dict={ x: data_to_test, y_: my_y_, keep_prob: 1.0})) print ("test accuracy %r"%y_conv.eval(feed_dict={ x: data_to_test, y_: my_y_, keep_prob: 1.0})) ''' ''' MNIST测试数据 print("The MNIST test data") print ("test accuracy %g"%accuracy.eval(feed_dict={ x: mnist.test.images, y_: mnist.test.labels, keep_prob: 1.0})) ''' M=0 N=0 for root,dirs,files in os.walk('C:/Users/Administrator/Desktop/ml_test/my_test_newdata'): for filespath in files: M=M+1 im = Image.open("C:/Users/Administrator/Desktop/ml_test/my_test_newdata/"+filespath) #转换灰度图 im_L =numpy.mat(im.convert("L")) im_dL=img_as_ubyte(im_L) im_dL=numpy.float32(im_dL/255.0) #data_to_test=cv2.resize(im_dL,(784,1)) data_to_test= numpy.mat(im_dL[0,:]) for i in range(1,28): data_to_test=numpy.hstack ((data_to_test,numpy.mat(im_dL[i,:]))) the_output= numpy.mat(y_conv.eval(feed_dict={x: data_to_test, y_: my_y_, keep_prob: 1.0})) the_out=numpy.max(the_output) for i in range(0,10): if the_out == the_output[0,i]: print("%s The digital is %d" %(filespath ,i)) if i == 1: N=N+1 print(N/M) ''' file_name='C:/Users/Administrator/Desktop/ml_test/NNT_binT/9/9_491.jpg'#导入自己的图片地址 im = Image.open(file_name).convert('L') tv = list(im.getdata()) #get pixel values tva = [ (255-x)*1.0/255.0 for x in tv] #print(tva) data_to_test=numpy.mat(tva) the_output= numpy.mat(y_conv.eval(feed_dict={x: data_to_test, y_: my_y_, keep_prob: 1.0})) the_out=numpy.max(the_output) for i in range(0,10): if the_out == the_output[0,i]: print("The digital is %d"%i) ''' ''' Num_AllFile=0 Num_RigFile=0 for root,dirs,files in os.walk('C:/Users/Administrator/Desktop/ml_test/NNT_binT/0'): for filespath in files: #print(filespath) Num_AllFile=Num_AllFile+1 im = Image.open("C:/Users/Administrator/Desktop/ml_test/NNT_binT/0/"+filespath) #转换灰度图 im_L =numpy.mat(im.convert("L")) im_dL=img_as_ubyte(im_L) im_dL=numpy.float32(im_dL/255.0) data_to_test= numpy.mat(im_dL[0,:]) for i in range(1,28): data_to_test=numpy.hstack ((data_to_test,numpy.mat(im_dL[i,:]))) the_output= numpy.mat(y_conv.eval(feed_dict={x: data_to_test, y_: my_y_, keep_prob: 1.0})) the_out=numpy.max(the_output) for i in range(0,10): if the_out == the_output[0,i]: #print("The digital is %d"%i) if i == 0: Num_RigFile=Num_RigFile+1 print("The accuracy of digital 9 is %f"%(Num_RigFile/Num_AllFile)) ''' ''' MNIST测试数据 ''' ''' for The_num in range(0,100): the_out=numpy.max(mnist.test.labels[The_num]) for i in range(0,10): if the_out == mnist.test.labels[The_num,i]: print(i) In_data=numpy.mat(mnist.test.images[The_num,:]) In_data=In_data*255 Input_data= numpy.mat(In_data[0,0:28]) for i in range(1,28): Input_data=numpy.vstack((Input_data,numpy.mat(In_data[0,28*i:(28*(i+1))]))) im_dL=numpy.float32(Input_data/255.0) data_to_test= numpy.mat(im_dL[0,:]) for i in range(1,28): data_to_test=numpy.hstack ((data_to_test,numpy.mat(im_dL[i,:]))) the_output= numpy.mat(y_conv.eval(feed_dict={x: data_to_test, y_: my_y_, keep_prob: 1.0})) the_out=numpy.max(the_output) for i in range(0,10): if the_out == the_output[0,i]: print("The digital is %d"%i) ''' ''' MNIST测试数据 print("The MNIST test data") print ("test accuracy %g"%accuracy.eval(feed_dict={ x: mnist.test.images, y_: mnist.test.labels, keep_prob: 1.0})) ''' """ ''' 保存结果 ''' saver = tf.train.Saver({'W_conv1':W_conv1,'b_conv1':b_conv1,'W_conv2':W_conv2,'b_conv2':b_conv2,'W_fc1':W_fc1,'b_fc1':b_fc1,'W_fc2':W_fc2,'b_fc2':b_fc2}) saver.save(sess,'C:/Users/Administrator/Desktop/ml_test/Train_result/Result') """ ''' 读取结果 ''' ''' saver.restore(sess,'C:/Users/Administrator/Desktop/ml_test/Train_result/Result') ''' print