from skimage import io, transform # skimage模塊下的io transform(圖像的形變與縮放)模塊 import glob # glob 文件通配符模塊 import os # os 處理文件和目錄的模塊 import tensorflow as tf import numpy as np # 多維數據處理模塊 import time import matplotlib.pyplot as plt # 數據集地址 #path = './flower_photos/' path = '../dataset/train/' # 模型保存地址 model_path = './cnn1_model.ckpt'# 將所有的圖片resize成100*100 w = 100 h = 100 c = 3# 讀取圖片+數據處理 def read_img(path):# os.listdir(path) 返回path指定的文件夾包含的文件或文件夾的名字的列表# os.path.isdir(path)判斷path是否是目錄# b = [x+x for x in list1 if x+x<15 ] 列表生成式,循環list1，當if為真時，將x+x加入列表bcate = [path + x for x in os.listdir(path) if os.path.isdir(path + x)]imgs = []labels = []print("開始讀入圖片和標簽。。。。")for idx, folder in enumerate(cate):# glob.glob(s+'*.py') 從目錄通配符搜索中生成文件列表for im in glob.glob(folder + '/*.png'):# 輸出讀取的圖片的名稱#print('reading the images:%s' % (im))# io.imread(im)讀取單張RGB圖片 skimage.io.imread(fname,as_grey=True)讀取單張灰度圖片# 讀取的圖片img = io.imread(im)# skimage.transform.resize(image, output_shape)改變圖片的尺寸img = transform.resize(img, (w, h))# 將讀取的圖片數據加載到imgs[]列表中imgs.append(img)# 將圖片的label加載到labels[]中，與上方的imgs索引對應labels.append(idx)# 將讀取的圖片和labels信息，轉化為numpy結構的ndarr(N維數組對象（矩陣）)數據信息print("讀入圖片和標簽完畢。。。。")return np.asarray(imgs, np.float32), np.asarray(labels, np.int32)# 調用讀取圖片的函數，得到圖片和labels的數據集 data, label = read_img(path)# 打亂順序 # 讀取data矩陣的第一維數（圖片的個數） num_example = data.shape[0] # 產生一個num_example范圍，步長為1的序列 arr = np.arange(num_example) # 調用函數，打亂順序 np.random.shuffle(arr) # 按照打亂的順序，重新排序 data = data[arr] label = label[arr]# 將所有數據分為訓練集和驗證集 ratio = 0.8 s = np.int(num_example * ratio) x_train = data[:s] y_train = label[:s] x_val = data[s:] y_val = label[s:]# -----------------構建網絡---------------------- # 本程序cnn網絡模型，共有7層，前三層為卷積層，后三層為全連接層，前三層中，每層包含卷積、激活、池化層 # 占位符設置輸入參數的大小和格式 x = tf.placeholder(tf.float32, shape=[None, w, h, c], name='x') y_ = tf.placeholder(tf.int32, shape=[None, ], name='y_')def inference(input_tensor, train, regularizer):# -----------------------第一層----------------------------with tf.variable_scope('layer1-conv1'):# 初始化權重conv1_weights為可保存變量，大小為5x5,3個通道（RGB），數量為32個conv1_weights = tf.get_variable("weight", [5, 5, 3, 32],initializer=tf.truncated_normal_initializer(stddev=0.1))# 初始化偏置conv1_biases，數量為32個conv1_biases = tf.get_variable("bias", [32], initializer=tf.constant_initializer(0.0))# 卷積計算，tf.nn.conv2d為tensorflow自帶2維卷積函數，input_tensor為輸入數據，# conv1_weights為權重，strides=[1, 1, 1, 1]表示左右上下滑動步長為1，padding='SAME'表示輸入和輸出大小一樣，即補0conv1 = tf.nn.conv2d(input_tensor, conv1_weights, strides=[1, 1, 1, 1], padding='SAME')# 激勵計算，調用tensorflow的relu函數relu1 = tf.nn.relu(tf.nn.bias_add(conv1, conv1_biases))with tf.name_scope("layer2-pool1"):# 池化計算，調用tensorflow的max_pool函數，strides=[1,2,2,1]，表示池化邊界，2個對一個生成，padding="VALID"表示不操作。pool1 = tf.nn.max_pool(relu1, ksize=[1, 2, 2, 1], strides=[1, 2, 2, 1], padding="VALID")# -----------------------第二層----------------------------with tf.variable_scope("layer3-conv2"):# 同上，不過參數的有變化，根據卷積計算和通道數量的變化，設置對應的參數conv2_weights = tf.get_variable("weight", [5, 5, 32, 64],initializer=tf.truncated_normal_initializer(stddev=0.1))conv2_biases = tf.get_variable("bias", [64], initializer=tf.constant_initializer(0.0))conv2 = tf.nn.conv2d(pool1, conv2_weights, strides=[1, 1, 1, 1], padding='SAME')relu2 = tf.nn.relu(tf.nn.bias_add(conv2, conv2_biases))with tf.name_scope("layer4-pool2"):pool2 = tf.nn.max_pool(relu2, ksize=[1, 2, 2, 1], strides=[1, 2, 2, 1], padding='VALID')# -----------------------第三層----------------------------# 同上，不過參數的有變化，根據卷積計算和通道數量的變化，設置對應的參數with tf.variable_scope("layer5-conv3"):conv3_weights = tf.get_variable("weight", [3, 3, 64, 128],initializer=tf.truncated_normal_initializer(stddev=0.1))conv3_biases = tf.get_variable("bias", [128], initializer=tf.constant_initializer(0.0))conv3 = tf.nn.conv2d(pool2, conv3_weights, strides=[1, 1, 1, 1], padding='SAME')relu3 = tf.nn.relu(tf.nn.bias_add(conv3, conv3_biases))with tf.name_scope("layer6-pool3"):pool3 = tf.nn.max_pool(relu3, ksize=[1, 2, 2, 1], strides=[1, 2, 2, 1], padding='VALID')# -----------------------第四層----------------------------# 同上，不過參數的有變化，根據卷積計算和通道數量的變化，設置對應的參數with tf.variable_scope("layer7-conv4"):conv4_weights = tf.get_variable("weight", [3, 3, 128, 128],initializer=tf.truncated_normal_initializer(stddev=0.1))conv4_biases = tf.get_variable("bias", [128], initializer=tf.constant_initializer(0.0))conv4 = tf.nn.conv2d(pool3, conv4_weights, strides=[1, 1, 1, 1], padding='SAME')relu4 = tf.nn.relu(tf.nn.bias_add(conv4, conv4_biases))with tf.name_scope("layer8-pool4"):pool4 = tf.nn.max_pool(relu4, ksize=[1, 2, 2, 1], strides=[1, 2, 2, 1], padding='VALID')nodes = 6 * 6 * 128reshaped = tf.reshape(pool4, [-1, nodes])# 使用變形函數轉化結構# -----------------------第五層---------------------------with tf.variable_scope('layer9-fc1'):# 初始化全連接層的參數，隱含節點為1024個fc1_weights = tf.get_variable("weight", [nodes, 1024],initializer=tf.truncated_normal_initializer(stddev=0.1))if regularizer != None: tf.add_to_collection('losses', regularizer(fc1_weights)) # 正則化矩陣fc1_biases = tf.get_variable("bias", [1024], initializer=tf.constant_initializer(0.1))# 使用relu函數作為激活函數fc1 = tf.nn.relu(tf.matmul(reshaped, fc1_weights) + fc1_biases)# 采用dropout層，減少過擬合和欠擬合的程度，保存模型最好的預測效率if train: fc1 = tf.nn.dropout(fc1, 0.5)# -----------------------第六層----------------------------with tf.variable_scope('layer10-fc2'):# 同上，不過參數的有變化，根據卷積計算和通道數量的變化，設置對應的參數fc2_weights = tf.get_variable("weight", [1024, 512],initializer=tf.truncated_normal_initializer(stddev=0.1))if regularizer != None: tf.add_to_collection('losses', regularizer(fc2_weights))fc2_biases = tf.get_variable("bias", [512], initializer=tf.constant_initializer(0.1))fc2 = tf.nn.relu(tf.matmul(fc1, fc2_weights) + fc2_biases)if train: fc2 = tf.nn.dropout(fc2, 0.5)# -----------------------第七層----------------------------with tf.variable_scope('layer11-fc3'):# 同上，不過參數的有變化，根據卷積計算和通道數量的變化，設置對應的參數fc3_weights = tf.get_variable("weight", [512, 5],initializer=tf.truncated_normal_initializer(stddev=0.1))if regularizer != None: tf.add_to_collection('losses', regularizer(fc3_weights))fc3_biases = tf.get_variable("bias", [5], initializer=tf.constant_initializer(0.1))logit = tf.matmul(fc2, fc3_weights) + fc3_biases # matmul矩陣相乘# 返回最后的計算結果return logit# ---------------------------網絡結束--------------------------- # 設置正則化參數為0.0001 regularizer = tf.contrib.layers.l2_regularizer(0.0001) # 將上述構建網絡結構引入 logits = inference(x, False, regularizer)# (小處理)將logits乘以1賦值給logits_eval，定義name，方便在后續調用模型時通過tensor名字調用輸出tensor b = tf.constant(value=1, dtype=tf.float32) logits_eval = tf.multiply(logits, b, name='logits_eval') # b為1# 設置損失函數，作為模型訓練優化的參考標準，loss越小，模型越優 loss = tf.nn.sparse_softmax_cross_entropy_with_logits(logits=logits, labels=y_) # 設置整體學習率為α為0.001 train_op = tf.train.AdamOptimizer(learning_rate=0.001).minimize(loss) # 設置預測精度 correct_prediction = tf.equal(tf.cast(tf.argmax(logits, 1), tf.int32), y_) acc = tf.reduce_mean(tf.cast(correct_prediction, tf.float32))# 定義一個函數，按批次取數據 def minibatches(inputs=None, targets=None, batch_size=None, shuffle=False):assert len(inputs) == len(targets)if shuffle:indices = np.arange(len(inputs))np.random.shuffle(indices)for start_idx in range(0, len(inputs) - batch_size + 1, batch_size): #range(start,end,step)if shuffle:excerpt = indices[start_idx:start_idx + batch_size]else:excerpt = slice(start_idx, start_idx + batch_size)yield inputs[excerpt], targets[excerpt]# 訓練和測試數據，可將n_epoch設置更大一些# 迭代次數 n_epoch = 30 fig_loss = np.zeros([n_epoch]) fig_acc1 = np.zeros([n_epoch]) fig_acc2= np.zeros([n_epoch]) # 每次迭代輸入的圖片數據 batch_size = 64 saver = tf.train.Saver(max_to_keep=1) # 可以指定保存的模型個數，利用max_to_keep=4，則最終會保存4個模型（ with tf.Session() as sess:# 初始化全局參數sess.run(tf.global_variables_initializer())# 開始迭代訓練，調用的都是前面設置好的函數或變量for epoch in range(n_epoch):start_time = time.time()# training#訓練集train_loss, train_acc, n_batch = 0, 0, 0for x_train_a, y_train_a in minibatches(x_train, y_train, batch_size, shuffle=True):_, err, ac = sess.run([train_op, loss, acc], feed_dict={x: x_train_a, y_: y_train_a})train_loss += errtrain_acc += acn_batch += 1if n_batch%20==0:# print("Epoch:%d After %d batch_size train loss" % (n_epoch,n_batch))# print(err)print("Epoch:%d After %d batch_size average train loss: %f" % (epoch, n_batch, np.sum(train_loss) / n_batch))# print("Epoch:%d After %d batch_size train acc %f" % (epoch, n_batch,ac))print("Epoch:%d After %d batch_size average train acc: %f" % (epoch, n_batch, np.sum(train_acc) / n_batch))#Epoch: 9 After 45 batch_size average train loss: 2.750402 Epoch: 9#After 45 batch_size average train acc: 0.993403fig_loss[epoch] = np.sum(train_loss) / n_batchfig_acc1[epoch] = np.sum(train_acc) / n_batch#validation#驗證集val_loss, val_acc, n_batch = 0, 0, 0for x_val_a, y_val_a in minibatches(x_val, y_val, batch_size, shuffle=False):err, ac = sess.run([loss, acc], feed_dict={x: x_val_a, y_: y_val_a})val_loss += errval_acc += acn_batch += 1print("validation loss: %f" % (np.sum(val_loss) / n_batch))print("validation acc: %f" % (np.sum(val_acc) / n_batch))fig_acc2[epoch] = np.sum(val_acc) / n_batch#保存模型及模型參數if epoch % 2 == 0:saver.save(sess, model_path, global_step=epoch)# 訓練loss圖 fig, ax1 = plt.subplots() lns1 = ax1.plot(np.arange(n_epoch), fig_loss, label="Loss") ax1.set_xlabel('iteration') ax1.set_ylabel('training loss')# 訓練和驗證兩種準確率曲線圖放在一張圖中 fig2, ax2 = plt.subplots() ax3 = ax2.twinx()#由ax2圖生成ax3圖 lns2 = ax2.plot(np.arange(n_epoch), fig_acc1, label="Loss") lns3 = ax3.plot(np.arange(n_epoch), fig_acc2, label="Loss")ax2.set_xlabel('iteration') ax2.set_ylabel('training acc') ax3.set_ylabel('val acc')# 合并圖例 lns = lns3 + lns2 labels = ["train acc", "val acc"] plt.legend(lns, labels, loc=7)plt.show()

?

總結

以上是生活随笔為你收集整理的我的模型--的全部內容，希望文章能夠幫你解決所遇到的問題。

如果覺得生活随笔網站內容還不錯，歡迎將生活随笔推薦給好友。