tensorflow（六）训练分类自己的图片（CNN超详细入门版）

之前一直用caffe做图像方面的东西，由于tensorflow环境配置简单，综合表现较为出色，因此打算转战tensorflow。学习这个框架，最开始还是要跑一跑文档中的mnist小程序（具体请参照tensorflow官方文档）。但是mnist中都是处理好的数据，具体的数据处理过程又没有讲，如果想要处理自己的图片数据，便有些无从下手，直接看源码的话又比较枯燥晦涩。这里是一份从图片的预处理到最终测试一张单独图片的完整的代码，供大家参考。由于一些原因，数据集不能发到网上，大概就是一个具有五种类型图片的数据集。

这里我把整个工程文件放上来：https://pan.baidu.com/s/1SSB8U2-DIqmUsgsI0BRVVw
其中log下是训练好的模型，可以直接运行程序，如果想要自己训练模型，可以把log文件夹删掉，然后讲代码最后测试图片的部分注释掉，再在最后加上一行run_training()调用训练函数即可。

这里由于只有很少的图片，很少的训练次数，因此测试结果不是很理想，大家可以增加大量的数据和增加训练次数来改善测试结果。

（PS:需要百度内推的可以加我好友~ v:f1979446971）

一：数据预处理

import os
import numpy as np
from PIL import Image
import tensorflow as tf
import matplotlib.pyplot as plt
#导入必要的包

train_dir = 'D:/picture/train/'
#存放用来训练的图片的路径

def get_files(file_dir):
    A5 = []
    label_A5 = []
    A6 = []
    label_A6 = []
    SEG = []
    label_SEG = []
    SUM = []
    label_SUM = []
    LTAX1 = []
    label_LTAX1 = []
    #定义存放各类别数据和对应标签的列表，列表名对应你所需要分类的列别名
    #A5，A6等是我的数据集中要分类图片的名字


    for file in os.listdir(file_dir):
        name = file.split(sep='.')
        if name[0]=='A5':
            A5.append(file_dir+file)
            label_A5.append(0)
        elif name[0] == 'A6':
            A6.append(file_dir+file)
            label_A6.append(1)
        elif name[0]=='LTAX1':
            LTAX1.append(file_dir+file)
            label_LTAX1.append(2)
        elif name[0] == 'SEG':
            SEG.append(file_dir+file)
            label_SEG.append(3)
        else:
            SUM.append(file_dir+file)
            label_SUM.append(4)
       #根据图片的名称，对图片进行提取，这里用.来进行划分
       ###这里一定要注意，如果是多分类问题的话，一定要将分类的标签从0开始。这里是五类，标签为0，1，2，3，4。我之前以为这个标签应该是随便设置的，结果就出现了Target[0] out of range的错误。
       
    print('There are %d A5\nThere are %d A6\nThere are %d LTAX1\nThere are %d SEG\nThere are %d SUM' \
          %(len(A5),len(A6),len(LTAX1),len(SEG),len(SUM)))
   #打印出提取图片的情况，检测是否正确提取

    image_list = np.hstack((A5,A6,LTAX1,SEG,SUM))
    label_list = np.hstack((label_A5,label_A6,label_LTAX1,label_SEG,label_SUM))
    #用来水平合并数组
   
    temp = np.array([image_list,label_list])
    temp = temp.transpose()
    np.random.shuffle(temp)
    
    image_list = list(temp[:,0])
    label_list = list(temp[:,1])
    label_list = [int(i) for i in label_list]
  
    return  image_list,label_list
    #返回两个list

def get_batch(image,label,image_W,image_H,batch_size,capacity):
    image = tf.cast(image,tf.string)
    label = tf.cast(label,tf.int32)
    #tf.cast()用来做类型转换

    input_queue = tf.train.slice_input_producer([image,label])
    #加入队列
    
    label = input_queue[1]
    image_contents = tf.read_file(input_queue[0])
    image = tf.image.decode_jpeg(image_contents,channels=3)
    #jpeg或者jpg格式都用decode_jpeg函数，其他格式可以去查看官方文档
    
    image = tf.image.resize_image_with_crop_or_pad(image,image_W,image_H)
    #resize
    
    image = tf.image.per_image_standardization(image)
    #对resize后的图片进行标准化处理
    
    image_batch,label_batch = tf.train.batch([image,label],batch_size = batch_size,num_threads=16,capacity = capacity)
    
    label_batch = tf.reshape(label_batch,[batch_size])
    return image_batch,label_batch
    #获取两个batch，两个batch即为传入神经网络的数据

对预处理的数据进行可视化，查看预处理的效果

BATCH_SIZE = 5
CAPACITY = 64
IMG_W = 208
IMG_H = 208

train_dir = 'D:/picture/train/'

image_list,label_list = get_files(train_dir)
image_batch,label_batch = get_batch(image_list,label_list,IMG_W,IMG_H,BATCH_SIZE,CAPACITY)

with tf.Session() as sess:
    i=0
    coord = tf.train.Coordinator()
    threads = tf.train.start_queue_runners(coord = coord)
    try:
        while not coord.should_stop() and i<2:
        #提取出两个batch的图片并可视化。
            img,label = sess.run([image_batch,label_batch])
            
            for j in np.arange(BATCH_SIZE):
                print('label: %d'%label[j])
                plt.imshow(img[j,:,:,:])
                plt.show()
            i+=1
    except tf.errors.OutOfRangeError:
        print('done!')
    finally:
        coord.request_stop()
    coord.join(threads)

**

二、设计神经网络模型

在设计神经网络的过程中，一定要对每一层的数据流动比较了解，弄清楚图片size的变化，不然会报错。
在进行测试网络模型的过程中，如果用的是IPython的话，要经常重新启动kernel，不然会出现conv1等层scope已经定义的错误。刚开始的时候这个问题困扰了很久，以为是定义变量作用域的过程中，语法使用错误，后来才知道是需要重新启动kernel。具体其中的原因我也不太清楚。
**

def inference(images, batch_size, n_classes):
    # conv1, shape = [kernel_size, kernel_size, channels, kernel_numbers]
    with tf.variable_scope("conv1") as scope:
        weights = tf.get_variable("weights",
                                  shape=[3, 3, 3, 16],
                                  dtype=tf.float32,
                                  initializer=tf.truncated_normal_initializer(stddev=0.1, dtype=tf.float32))
        biases = tf.get_variable("biases",
                                 shape=[16],
                                 dtype=tf.float32,
                                 initializer=tf.constant_initializer(0.1))
        conv = tf.nn.conv2d(images, weights, strides=[1, 1, 1, 1], padding="SAME")
        pre_activation = tf.nn.bias_add(conv, biases)
        conv1 = tf.nn.relu(pre_activation, name="conv1")

    # pool1 && norm1
    with tf.variable_scope("pooling1_lrn") as scope:
        pool1 = tf.nn.max_pool(conv1, ksize=[1, 3, 3, 1], strides=[1, 2, 2, 1],
                               padding="SAME", name="pooling1")
        norm1 = tf.nn.lrn(pool1, depth_radius=4, bias=1.0, alpha=0.001/9.0,
                          beta=0.75, name='norm1')

    # conv2
    with tf.variable_scope("conv2") as scope:
        weights = tf.get_variable("weights",
                                  shape=[3, 3, 16, 16],
                                  dtype=tf.float32,
                                  initializer=tf.truncated_normal_initializer(stddev=0.1, dtype=tf.float32))
        biases = tf.get_variable("biases",
                                 shape=[16],
                                 dtype=tf.float32,
                                 initializer=tf.constant_initializer(0.1))
        conv = tf.nn.conv2d(norm1, weights, strides=[1, 1, 1, 1], padding="SAME")
        pre_activation = tf.nn.bias_add(conv, biases)
        conv2 = tf.nn.relu(pre_activation, name="conv2")

    # pool2 && norm2
    with tf.variable_scope("pooling2_lrn") as scope:
        pool2 = tf.nn.max_pool(conv2, ksize=[1, 3, 3, 1], strides=[1, 2, 2, 1],
                               padding="SAME", name="pooling2")
        norm2 = tf.nn.lrn(pool2, depth_radius=4, bias=1.0, alpha=0.001/9.0,
                          beta=0.75, name='norm2')

    # full-connect1
    with tf.variable_scope("fc1") as scope:
        reshape = tf.reshape(norm2, shape=[batch_size, -1])
        dim = reshape.get_shape()[1].value
        weights = tf.get_variable("weights",
                                  shape=[dim, 128],
                                  dtype=tf.float32,
                                  initializer=tf.truncated_normal_initializer(stddev=0.005, dtype=tf.float32))
        biases = tf.get_variable("biases",
                                 shape=[128],
                                 dtype=tf.float32,
                                 initializer=tf.constant_initializer(0.1))
        fc1 = tf.nn.relu(tf.matmul(reshape, weights) + biases, name="fc1")

    # full_connect2
    with tf.variable_scope("fc2") as scope:
        weights = tf.get_variable("weights",
                                  shape=[128, 128],
                                  dtype=tf.float32,
                                  initializer=tf.truncated_normal_initializer(stddev=0.005, dtype=tf.float32))
        biases = tf.get_variable("biases",
                                 shape=[128],
                                 dtype=tf.float32,
                                 initializer=tf.constant_initializer(0.1))
        fc2 = tf.nn.relu(tf.matmul(fc1, weights) + biases, name="fc2")

    # softmax
    with tf.variable_scope("softmax_linear") as scope:
        weights = tf.get_variable("weights",
                                  shape=[128, n_classes],
                                  dtype=tf.float32,
                                  initializer=tf.truncated_normal_initializer(stddev=0.005, dtype=tf.float32))
        biases = tf.get_variable("biases",
                                 shape=[n_classes],
                                 dtype=tf.float32,
                                 initializer=tf.constant_initializer(0.1))
        softmax_linear = tf.add(tf.matmul(fc2, weights), biases, name="softmax_linear")
    return softmax_linear

def losses(logits, labels):
    with tf.variable_scope("loss") as scope:
        cross_entropy = tf.nn.sparse_softmax_cross_entropy_with_logits(logits=logits,
                                                                       labels=labels, name="xentropy_per_example")
        loss = tf.reduce_mean(cross_entropy, name="loss")
        tf.summary.scalar(scope.name + "loss", loss)
    return loss

def trainning(loss, learning_rate):
    with tf.name_scope("optimizer"):
        optimizer = tf.train.AdamOptimizer(learning_rate=learning_rate)
        global_step = tf.Variable(0, name="global_step", trainable=False)
        train_op = optimizer.minimize(loss, global_step=global_step)
    return train_op

def evaluation(logits, labels):
    with tf.variable_scope("accuracy") as scope:
        correct = tf.nn.in_top_k(logits, labels, 1)
        correct = tf.cast(correct, tf.float16)
        accuracy = tf.reduce_mean(correct)
        tf.summary.scalar(scope.name + "accuracy", accuracy)
    return accuracy

N_CLASSES = 5
#要分类的类别数，这里是5分类
IMG_W = 208
IMG_H = 208
#设置图片的size
BATCH_SIZE = 8
CAPACITY = 64
MAX_STEP = 1000
#迭代一千次，如果机器配置好的话，建议至少10000次以上
learning_rate = 0.0001
#学习率

**

三、训练

**

def run_training():
    train_dir = 'D:/picture/train/'
    logs_train_dir = 'D:/picture/log/'
    #存放一些模型文件的目录
    train,train_label = get_files(train_dir)
    train_batch,train_label_batch = get_batch(train,train_label,
                                                         IMG_W,
                                                         IMG_H,
                                                         BATCH_SIZE,
                                                         CAPACITY)
    train_logits =inference(train_batch,BATCH_SIZE,N_CLASSES)
    train_loss = losses(train_logits,train_label_batch)
    train_op = trainning(train_loss,learning_rate)
    train_acc = evaluation(train_logits,train_label_batch)
    
    summary_op = tf.summary.merge_all()
    sess = tf.Session()
    train_writer = tf.summary.FileWriter(logs_train_dir,sess.graph)
    saver = tf.train.Saver()
    
    sess.run(tf.global_variables_initializer())
    coord = tf.train.Coordinator()
    threads = tf.train.start_queue_runners(sess = sess,coord = coord)
    
    try:
        for step in np.arange(MAX_STEP):
            if coord.should_stop():
                break
            _,tra_loss,tra_acc = sess.run([train_op,train_loss,train_acc])
            if step %  50 == 0:
                print('Step %d,train loss = %.2f,train occuracy = %.2f%%'%(step,tra_loss,tra_acc))
                #每迭代50次，打印出一次结果
                summary_str = sess.run(summary_op)
                train_writer.add_summary(summary_str,step)
                
            if step % 200 ==0 or (step +1) == MAX_STEP:
                checkpoint_path = os.path.join(logs_train_dir,'model.ckpt')
                saver.save(sess,checkpoint_path,global_step = step)
                #每迭代200次，利用saver.save()保存一次模型文件，以便测试的时候使用
                
    except tf.errors.OutOfRangeError:
        print('Done training epoch limit reached')
    finally:
        coord.request_stop()
    
    coord.join(threads)
    sess.close()

**

四、测试一张图片

**

def get_one_image(img_dir):
     image = Image.open(img_dir)
     #Image.open()
     #好像一次只能打开一张图片，不能一次打开一个文件夹，这里大家可以去搜索一下
     plt.imshow(image)
     image = image.resize([208, 208])
     image_arr = np.array(image)
     return image_arr

def test(test_file):
    log_dir = 'D:/picture/log/'
    image_arr = get_one_image(test_file)
    
    with tf.Graph().as_default():
        image = tf.cast(image_arr, tf.float32)
        image = tf.image.per_image_standardization(image)
        image = tf.reshape(image, [1,208, 208, 3])
        print(image.shape)
        p = inference(image,1,5)
        logits = tf.nn.softmax(p)
        x = tf.placeholder(tf.float32,shape = [208,208,3])
        saver = tf.train.Saver()
        with tf.Session() as sess:
            ckpt = tf.train.get_checkpoint_state(log_dir)
            if ckpt and ckpt.model_checkpoint_path:
                global_step = ckpt.model_checkpoint_path.split('/')[-1].split('-')[-1]
                saver.restore(sess, ckpt.model_checkpoint_path)
                #调用saver.restore()函数，加载训练好的网络模型
                
                print('Loading success')
            else:
                print('No checkpoint')
            prediction = sess.run(logits, feed_dict={x: image_arr})
            max_index = np.argmax(prediction) 
            print('预测的标签为：')
            print(max_index)
            print('预测的结果为：')
            print(prediction)
            
            if max_index==0:
                print('This is a LTAX with possibility %.6f' %prediction[:, 0])
            elif max_index == 1:
                print('This is a SUM with possibility %.6f' %prediction[:, 1])
            elif max_index == 2:
                print('This is a A5 with possibility %.6f' %prediction[:, 2])
            elif max_index == 3:
                print('This is a A6 with possibility %.6f' %prediction[:, 3])
            else :
                print('This is a SEG with possibility %.6f' %prediction[:, 4])

调用test函数测试图片的预测结果。

test('D:\\picture\\test\\A51.jpeg')
test('D:\\picture\\test\\A52.jpeg')
test('D:\\picture\\test\\A61.jpeg')
test('D:\\picture\\test\\A62.jpeg')
test('D:\\picture\\test\\LTAX1.jpeg')
test('D:\\picture\\test\\LTAX2.jpeg')
test('D:\\picture\\test\\SEG1.jpg')
test('D:\\picture\\test\\SEG2.jpg')
test('D:\\picture\\test\\SUM1.jpeg')
test('D:\\picture\\test\\SUM2.jpeg')