Unet_Own.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "from __future__ import print_function\n",
    "import numpy as np \n",
    "import os\n",
    "import skimage.io as io\n",
    "import skimage.transform as trans\n",
    "import numpy as np\n",
    "from keras.models import *\n",
    "from keras.layers import *\n",
    "from keras.optimizers import *\n",
    "from keras.callbacks import ModelCheckpoint, LearningRateScheduler\n",
    "from keras import backend as keras\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# For the imbalance of prospect and background, use dice loss\n",
    "def dice_coef(y_true, y_pred, smooth = 1):\n",
    "    y_true_f = K.flatten(y_true)\n",
    "    y_pred_f = K.flatten(y_pred)\n",
    "    intersection = K.sum(y_true_f * y_pred_f)\n",
    "    return (2. * intersection + smooth) / (K.sum(y_true_f * y_true_f) + K.sum(y_pred_f * y_pred_f) + smooth)\n",
    "\n",
    "# def dice_loss(y_true, y_pred):\n",
    "#     return 1. - dice_coef(y_true, y_pred)\n",
    "\n",
    "\n",
    "def unet(pretrained_weights = None,input_size = (256,256,1)):\n",
    "    inputs = Input(input_size)\n",
    "    conv1 = Conv2D(64, 3, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(inputs)\n",
    "    conv1=BatchNormalization()(conv1)\n",
    "    conv1 = Conv2D(64, 3, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(conv1)\n",
    "    conv1=BatchNormalization()(conv1)\n",
    "    pool1 = MaxPooling2D(pool_size=(2, 2))(conv1)\n",
    "    \n",
    "    \n",
    "    conv2 = Conv2D(128, 3, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(pool1)\n",
    "    conv2=BatchNormalization()(conv2)\n",
    "    conv2 = Conv2D(128, 3, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(conv2)\n",
    "    conv2=BatchNormalization()(conv2)\n",
    "    pool2 = MaxPooling2D(pool_size=(2, 2))(conv2)\n",
    "    \n",
    "    conv3 = Conv2D(256, 3, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(pool2)\n",
    "    conv3=BatchNormalization()(conv3)\n",
    "    conv3 = Conv2D(256, 3, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(conv3)\n",
    "    conv3=BatchNormalization()(conv3)\n",
    "    pool3 = MaxPooling2D(pool_size=(2, 2))(conv3)\n",
    "    \n",
    "    conv4 = Conv2D(512, 3, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(pool3)\n",
    "    conv4=BatchNormalization()(conv4)\n",
    "    conv4 = Conv2D(512, 3, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(conv4)\n",
    "    conv4=BatchNormalization()(conv4)\n",
    "    drop4 = Dropout(0.5)(conv4)\n",
    "    pool4 = MaxPooling2D(pool_size=(2, 2))(drop4)\n",
    "\n",
    "    conv5 = Conv2D(1024, 3, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(pool4)\n",
    "    conv5=BatchNormalization()(conv5)\n",
    "    conv5 = Conv2D(1024, 3, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(conv5)\n",
    "    conv5=BatchNormalization()(conv5)\n",
    "    drop5 = Dropout(0.5)(conv5)\n",
    "\n",
    "    up6 = Conv2D(512, 2, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(UpSampling2D(size = (2,2))(drop5))\n",
    "    conv6=BatchNormalization()(conv6)  \n",
    "    merge6 = concatenate([drop4,up6], axis = 3)\n",
    "    conv6 = Conv2D(512, 3, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(merge6)\n",
    "    conv6=BatchNormalization()(conv6)\n",
    "    conv6 = Conv2D(512, 3, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(conv6)\n",
    "    conv6=BatchNormalization()(conv6)\n",
    "\n",
    "    up7 = Conv2D(256, 2, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(UpSampling2D(size = (2,2))(conv6))\n",
    "    conv7=BatchNormalization()(conv7)\n",
    "    merge7 = concatenate([conv3,up7], axis = 3)\n",
    "    conv7 = Conv2D(256, 3, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(merge7)\n",
    "    conv7=BatchNormalization()(conv7)\n",
    "    conv7 = Conv2D(256, 3, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(conv7)\n",
    "    conv7=BatchNormalization()(conv7)\n",
    "\n",
    "    up8 = Conv2D(128, 2, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(UpSampling2D(size = (2,2))(conv7))\n",
    "    conv8=BatchNormalization()(conv8)\n",
    "    merge8 = concatenate([conv2,up8], axis = 3)\n",
    "    conv8 = Conv2D(128, 3, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(merge8)\n",
    "    conv8=BatchNormalization()(conv8)\n",
    "    conv8 = Conv2D(128, 3, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(conv8)\n",
    "    conv8=BatchNormalization()(conv8)\n",
    "\n",
    "    up9 = Conv2D(64, 2, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(UpSampling2D(size = (2,2))(conv8))\n",
    "    conv9=BatchNormalization()(conv9)\n",
    "    merge9 = concatenate([conv1,up9], axis = 3)\n",
    "    conv9 = Conv2D(64, 3, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(merge9)\n",
    "    conv9=BatchNormalization()(conv9)\n",
    "    conv9 = Conv2D(64, 3, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(conv9)\n",
    "    conv9=BatchNormalization()(conv9)\n",
    "    conv9 = Conv2D(2, 3, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(conv9)\n",
    "    conv9=BatchNormalization()(conv9)\n",
    "\n",
    "    conv10 = Conv2D(1, 1, activation = 'sigmoid')(conv9)\n",
    "\n",
    "    model = Model(input = inputs, output = conv10)\n",
    "\n",
    "    model.compile(optimizer = Adam(lr = 1e-3), loss = \"binary_crossentropy\", metrics = ['accuracy'])\n",
    "    \n",
    "    model.summary()\n",
    "\n",
    "    if(pretrained_weights):\n",
    "    \tmodel.load_weights(pretrained_weights)\n",
    "\n",
    "    return model\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(181, 512, 512, 1)\n",
      "(181, 512, 512, 1)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from keras.preprocessing.image import ImageDataGenerator\n",
    "import numpy as np \n",
    "import os\n",
    "from glob import glob\n",
    "import skimage.io as io\n",
    "import skimage.transform as trans\n",
    "import cv2\n",
    "\n",
    "\n",
    "img_size = (512,512)\n",
    "\n",
    "imgs=sorted(glob('./Prostate_Data/PCA_Seg/Image/*'))\n",
    "labels=sorted(glob('./Prostate_Data/PCA_Seg/Mask/*'))\n",
    "\n",
    "\n",
    "\n",
    "X_image = []\n",
    "Y_label = []\n",
    "\n",
    "for data in zip(imgs,labels):\n",
    "    img=cv2.imread(data[0],0)\n",
    "    img=cv2.resize(img,img_size)\n",
    "    img = img.astype('float32') / 255.0\n",
    "    mask=cv2.imread(data[1],0)\n",
    "    mask=cv2.resize(mask,img_size)\n",
    "    mask = mask.astype('float32') / 255.0\n",
    "    mask[mask > 0.5] = 1\n",
    "    mask[mask <= 0.5] = 0\n",
    "    \n",
    "\n",
    "    \n",
    "    X_image.append(img)\n",
    "    Y_label.append(mask)\n",
    "    \n",
    "x_train=np.array(X_image).reshape((len(X_image),img_size[0],img_size[1],1))\n",
    "y_train=np.array(Y_label).reshape((len(Y_label),img_size[0],img_size[1],1))\n",
    "\n",
    "\n",
    "print (x_train.shape)\n",
    "print (y_train.shape)\n",
    "\n",
    "\n",
    "plt.imshow(x_train[0,:,:,0],cmap=plt.cm.gray)\n",
    "plt.show()\n",
    "plt.imshow(y_train[0,:,:,0],cmap=plt.cm.gray)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ilab/anaconda2/envs/Dinghow/lib/python2.7/site-packages/ipykernel_launcher.py:89: UserWarning: Update your `Model` call to the Keras 2 API: `Model(outputs=Tensor(\"co..., inputs=Tensor(\"in...)`\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "__________________________________________________________________________________________________\n",
      "Layer (type)                    Output Shape         Param #     Connected to                     \n",
      "==================================================================================================\n",
      "input_1 (InputLayer)            (None, 512, 512, 1)  0                                            \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_1 (Conv2D)               (None, 512, 512, 64) 640         input_1[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_1 (BatchNor (None, 512, 512, 64) 256         conv2d_1[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_2 (Conv2D)               (None, 512, 512, 64) 36928       batch_normalization_1[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_2 (BatchNor (None, 512, 512, 64) 256         conv2d_2[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling2d_1 (MaxPooling2D)  (None, 256, 256, 64) 0           batch_normalization_2[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_3 (Conv2D)               (None, 256, 256, 128 73856       max_pooling2d_1[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_3 (BatchNor (None, 256, 256, 128 512         conv2d_3[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_4 (Conv2D)               (None, 256, 256, 128 147584      batch_normalization_3[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_4 (BatchNor (None, 256, 256, 128 512         conv2d_4[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling2d_2 (MaxPooling2D)  (None, 128, 128, 128 0           batch_normalization_4[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_5 (Conv2D)               (None, 128, 128, 256 295168      max_pooling2d_2[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_5 (BatchNor (None, 128, 128, 256 1024        conv2d_5[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_6 (Conv2D)               (None, 128, 128, 256 590080      batch_normalization_5[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_6 (BatchNor (None, 128, 128, 256 1024        conv2d_6[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling2d_3 (MaxPooling2D)  (None, 64, 64, 256)  0           batch_normalization_6[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_7 (Conv2D)               (None, 64, 64, 512)  1180160     max_pooling2d_3[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_7 (BatchNor (None, 64, 64, 512)  2048        conv2d_7[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_8 (Conv2D)               (None, 64, 64, 512)  2359808     batch_normalization_7[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_8 (BatchNor (None, 64, 64, 512)  2048        conv2d_8[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "dropout_1 (Dropout)             (None, 64, 64, 512)  0           batch_normalization_8[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling2d_4 (MaxPooling2D)  (None, 32, 32, 512)  0           dropout_1[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_9 (Conv2D)               (None, 32, 32, 1024) 4719616     max_pooling2d_4[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_9 (BatchNor (None, 32, 32, 1024) 4096        conv2d_9[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_10 (Conv2D)              (None, 32, 32, 1024) 9438208     batch_normalization_9[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_10 (BatchNo (None, 32, 32, 1024) 4096        conv2d_10[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "dropout_2 (Dropout)             (None, 32, 32, 1024) 0           batch_normalization_10[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "up_sampling2d_1 (UpSampling2D)  (None, 64, 64, 1024) 0           dropout_2[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_11 (Conv2D)              (None, 64, 64, 512)  2097664     up_sampling2d_1[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_1 (Concatenate)     (None, 64, 64, 1024) 0           dropout_1[0][0]                  \n",
      "                                                                 conv2d_11[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_12 (Conv2D)              (None, 64, 64, 512)  4719104     concatenate_1[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_13 (Conv2D)              (None, 64, 64, 512)  2359808     conv2d_12[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "up_sampling2d_2 (UpSampling2D)  (None, 128, 128, 512 0           conv2d_13[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_14 (Conv2D)              (None, 128, 128, 256 524544      up_sampling2d_2[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_2 (Concatenate)     (None, 128, 128, 512 0           batch_normalization_6[0][0]      \n",
      "                                                                 conv2d_14[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_15 (Conv2D)              (None, 128, 128, 256 1179904     concatenate_2[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_11 (BatchNo (None, 128, 128, 256 1024        conv2d_15[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_16 (Conv2D)              (None, 128, 128, 256 590080      batch_normalization_11[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_12 (BatchNo (None, 128, 128, 256 1024        conv2d_16[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "up_sampling2d_3 (UpSampling2D)  (None, 256, 256, 256 0           batch_normalization_12[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_17 (Conv2D)              (None, 256, 256, 128 131200      up_sampling2d_3[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_3 (Concatenate)     (None, 256, 256, 256 0           batch_normalization_4[0][0]      \n",
      "                                                                 conv2d_17[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_18 (Conv2D)              (None, 256, 256, 128 295040      concatenate_3[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_13 (BatchNo (None, 256, 256, 128 512         conv2d_18[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_19 (Conv2D)              (None, 256, 256, 128 147584      batch_normalization_13[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_14 (BatchNo (None, 256, 256, 128 512         conv2d_19[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "up_sampling2d_4 (UpSampling2D)  (None, 512, 512, 128 0           batch_normalization_14[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_20 (Conv2D)              (None, 512, 512, 64) 32832       up_sampling2d_4[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_4 (Concatenate)     (None, 512, 512, 128 0           batch_normalization_2[0][0]      \n",
      "                                                                 conv2d_20[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_21 (Conv2D)              (None, 512, 512, 64) 73792       concatenate_4[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_15 (BatchNo (None, 512, 512, 64) 256         conv2d_21[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_22 (Conv2D)              (None, 512, 512, 64) 36928       batch_normalization_15[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_16 (BatchNo (None, 512, 512, 64) 256         conv2d_22[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_23 (Conv2D)              (None, 512, 512, 2)  1154        batch_normalization_16[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_17 (BatchNo (None, 512, 512, 2)  8           conv2d_23[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_24 (Conv2D)              (None, 512, 512, 1)  3           batch_normalization_17[0][0]     \n",
      "==================================================================================================\n",
      "Total params: 31,051,149\n",
      "Trainable params: 31,041,417\n",
      "Non-trainable params: 9,732\n",
      "__________________________________________________________________________________________________\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 144 samples, validate on 37 samples\n",
      "Epoch 1/20\n",
      "144/144 [==============================] - 47s 326ms/step - loss: 0.6322 - acc: 0.9600 - val_loss: 2.3700 - val_acc: 0.2383\n",
      "Epoch 2/20\n",
      "144/144 [==============================] - 38s 263ms/step - loss: 0.5432 - acc: 0.9843 - val_loss: 0.7779 - val_acc: 0.8209\n",
      "Epoch 3/20\n",
      "144/144 [==============================] - 38s 264ms/step - loss: 0.4607 - acc: 0.9880 - val_loss: 1.6351 - val_acc: 0.8548\n",
      "Epoch 4/20\n",
      "144/144 [==============================] - 38s 264ms/step - loss: 0.3850 - acc: 0.9889 - val_loss: 0.3632 - val_acc: 0.9788\n",
      "Epoch 5/20\n",
      "144/144 [==============================] - 38s 265ms/step - loss: 0.3140 - acc: 0.9905 - val_loss: 0.6455 - val_acc: 0.9135\n",
      "Epoch 6/20\n",
      "144/144 [==============================] - 38s 265ms/step - loss: 0.2568 - acc: 0.9912 - val_loss: 0.4271 - val_acc: 0.9270\n",
      "Epoch 7/20\n",
      "144/144 [==============================] - 38s 265ms/step - loss: 0.2081 - acc: 0.9916 - val_loss: 0.2498 - val_acc: 0.9886\n",
      "Epoch 8/20\n",
      "144/144 [==============================] - 38s 266ms/step - loss: 0.1683 - acc: 0.9921 - val_loss: 0.1515 - val_acc: 0.9899\n",
      "Epoch 9/20\n",
      "144/144 [==============================] - 38s 267ms/step - loss: 0.1391 - acc: 0.9920 - val_loss: 0.1522 - val_acc: 0.9828\n",
      "Epoch 10/20\n",
      "144/144 [==============================] - 39s 268ms/step - loss: 0.1263 - acc: 0.9905 - val_loss: 2.4398 - val_acc: 0.3916\n",
      "Epoch 11/20\n",
      "144/144 [==============================] - 39s 268ms/step - loss: 0.1084 - acc: 0.9922 - val_loss: 0.5061 - val_acc: 0.9033\n",
      "Epoch 12/20\n",
      "144/144 [==============================] - 39s 268ms/step - loss: 0.0909 - acc: 0.9915 - val_loss: 0.2094 - val_acc: 0.9844\n",
      "Epoch 13/20\n",
      "144/144 [==============================] - 39s 268ms/step - loss: 0.0795 - acc: 0.9920 - val_loss: 0.0970 - val_acc: 0.9910\n",
      "Epoch 14/20\n",
      "144/144 [==============================] - 39s 268ms/step - loss: 0.0678 - acc: 0.9924 - val_loss: 0.1004 - val_acc: 0.9869\n",
      "Epoch 15/20\n",
      "144/144 [==============================] - 39s 268ms/step - loss: 0.0612 - acc: 0.9924 - val_loss: 0.0583 - val_acc: 0.9934\n",
      "Epoch 16/20\n",
      "144/144 [==============================] - 39s 268ms/step - loss: 0.0535 - acc: 0.9920 - val_loss: 0.0534 - val_acc: 0.9932\n",
      "Epoch 17/20\n",
      "144/144 [==============================] - 39s 268ms/step - loss: 0.0486 - acc: 0.9924 - val_loss: 0.0510 - val_acc: 0.9926\n",
      "Epoch 18/20\n",
      "144/144 [==============================] - 39s 268ms/step - loss: 0.0434 - acc: 0.9925 - val_loss: 0.0434 - val_acc: 0.9903\n",
      "Epoch 19/20\n",
      "144/144 [==============================] - 39s 269ms/step - loss: 0.0402 - acc: 0.9927 - val_loss: 0.0473 - val_acc: 0.9932\n",
      "Epoch 20/20\n",
      "144/144 [==============================] - 39s 268ms/step - loss: 0.0384 - acc: 0.9924 - val_loss: 0.0359 - val_acc: 0.9918\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x7f69c8f37d10>"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model=unet(input_size=(img_size[0],img_size[1],1))\n",
    "\n",
    "model.fit(x_train,y_train,epochs=20,batch_size=2,validation_split=0.2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1, 512, 512, 1)\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAQYAAAD8CAYAAACVSwr3AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvIxREBQAAFSVJREFUeJzt3WtonNedx/Hvf0Yazegu24l8kRLZ1KVNIJsGt03aEropXZJsqfMilJRCTQkY9gItXeg6u7BQ2DfdF01btrRrNmXdpZdkuy0xYXe7iR3YpZA0dnP1LVHSGkuRL7Et27IuM5r574s5z+xIR47HtkaakX4fGPQ8Z57R/BV5fjrnPOd5Yu6OiEi11HIXICKNR8EgIhEFg4hEFAwiElEwiEhEwSAikboEg5ndb2bHzGzYzHbV4z1EpH5ssdcxmFkaeBP4LDACvAR80d0PL+obiUjd1KPH8DFg2N3fcfc88HNgex3eR0TqpKUO33MTcKJqfwT4+Pu9wMy0/FKk/t5z95tqObAewVATM9sJ7Fyu9xdZhY7XemA9gmEUGKzaHwhtc7j7bmA3qMcg0mjqMcfwErDVzDabWQZ4BNhbh/cRkTpZ9B6Du8+a2V8CvwbSwI/c/dBiv4+I1M+in668riI0lBBZCgfdfVstB2rlo4hEFAwiElEwiEhEwSAiEQWDiEQUDCISUTCISETBICIRBYOIRBQMIhJRMIhIRMEgIhEFg4hEFAwiElEwiEhEwSAiEQWDiEQUDCISUTCISETBICIRBYOIRBQMIhJRMIhIRMEgIhEFg4hEFAwiElEwiEhEwSAiEQWDiEQUDCISUTCISETBICIRBYOIRBQMIhK5ajCY2Y/M7LSZvVHVtsbMnjWzt8LXvtBuZvY9Mxs2s9fM7K56Fi8i9VFLj+FfgPvnte0C9rn7VmBf2Ad4ANgaHjuBHyxOmSKylK4aDO7+P8C5ec3bgT1hew/wUFX7j73sBaDXzDYsVrEisjSud46h393HwvZJoD9sbwJOVB03EtoiZrbTzA6Y2YHrrEFE6qTlRr+Bu7uZ+XW8bjewG+B6Xi8i9XO9PYZTyRAhfD0d2keBwarjBkKbiDSR6w2GvcCOsL0DeLqq/cvh7MTdwIWqIYeINAt3f98H8DNgDChQnjN4FFhL+WzEW8BzwJpwrAHfB94GXge2Xe37h9e5HnroUffHgVo+j+6OhQ/mstIcg8iSOOju22o5UCsfRSSiYBCRiIJBRCIKBhGJKBhEJKJgEJGIgkFEIgoGEYkoGEQkomAQkYiCQUQiCgYRiSgYRCSiYBCRiIJBRCIKBhGJKBhEJKJgEJGIgkFEIgoGEYkoGEQkomAQkYiCQUQiCgYRiSgYRCSiYBCRiIJBRCIKBhGJKBhEJKJgEJGIgkFEIgoGEYkoGEQkctVgMLNBM3vezA6b2SEz+2poX2Nmz5rZW+FrX2g3M/uemQ2b2Wtmdle9fwgRWVy19Bhmgb9y99uAu4G/MLPbgF3APnffCuwL+wAPAFvDYyfwg0WvWkTq6qrB4O5j7v67sH0JOAJsArYDe8Jhe4CHwvZ24Mde9gLQa2YbFr1yEamba5pjMLMh4CPAi0C/u4+Fp04C/WF7E3Ci6mUjoU1EmkRLrQeaWSfw78DX3P2imVWec3c3M7+WNzaznZSHGiLSYGrqMZhZK+VQ+Im7/zI0n0qGCOHr6dA+CgxWvXwgtM3h7rvdfZu7b7ve4kWkPmo5K2HAE8ARd/921VN7gR1hewfwdFX7l8PZibuBC1VDDhFpAub+/iMAM/sU8L/A60ApNP8N5XmGp4BbgOPAF9z9XAiSfwTuByaBr7j7gau8xzUNQ0TkuhystYd+1WBYCgoGkSVRczBo5aOIRGo+KyGrm5lRfSaqVCq9z9HS7BQMckXVQZBKpaL92dnZ5ShLloCCQSLVAZD0FKp7DGZGKpUilUpRKBRohHkqWVwKhlWsOgAW2k/aqnsLSQiYGS0tLaRSKaanp+tfrCwpBcMqVP0XP9mH8oc+ecw/1szmzCskPYh0Ok1rayuFQmFpfwipKwXDKpRKpWhtbSWdTs/pDRSLRYrFIrOzs3N6BvMnHqEcIkl7S0sL7q45hxVEwbDKmBmtra1ks1laW1vJZDK0trbS0tJCoVBgamqKiYkJZmdnKZVKUY+huleRhEM6na4cl8/nl/tHlEWgYFhlkmBIQqGrq4vu7m66u7spFoucP3+esbExJiYmouFB9TAjCY0kHADS6fSS/zxSHwqGVST5657NZsnlcuRyOTZt2sTtt99OX18f6XSafD7P0aNHOXz4MO+++y5TU1OVEEi4+5z5hlQqVQkNzTesDFr5uIokwZDJZCrhcMstt7B+/Xo6OjpIp9P09fWxdetW7r33XtavXx8FwpUeyfPJe0hzUzCsIslEYTKM6OzsrIRC8lc+mXjs6+vjnnvuqUwsLrRWYaFwSN5noVOf0jw0lFglklBoa2ujvb2dzs5ObrrpJvr7++no6GDNmjX09/djZoyMjHDs2DE2b95MLpcjn8+/bzgk3z/ZT3oNOkvRvBQMq0QqlSKdTld6DJ2dnZVQaGtro6Ojg4GBAXK5HK2trZw9exZ3r6x1SM5IXOl7J8dUS6fTFIvF+v5gUhcKhlWgekFTcuoxm83S19c3JzDWrl1LZ2cnExMTdHZ2ks/nFwyEK62YTM5SJMcnk5K64Kr5aI5hhUulUrS0tFTWK5gZs7OzFIvFOb2H5MOdyWSYnp7m5ptv5tKlS+Tz+cqcQXWwLKR6viFZRJXJZDTf0ITUY1jBkrF+Lpejp6dnzoc6n88zMzNDPp+nUCiQyWQ4efIkZsbZs2cpFoucOnWKUqk0Z+m0u1MsFufMLVSflahuT4JkdnZW8w1NRsGwgqVSKXK5HBs3bmTbtm0Ui0VOnjzJxYsXKZVKnDlzhr6+PgqFAr29vYyNjTE9Pc2ZM2coFAocPny4soYhOQU5f9Vj0pZY6MrMdDpNqVTSkKKJKBhWsEwmw+bNm9m+fTu5XI4LFy5wxx13cPnyZS5dukSxWGR4eJju7m42bNhAe3s7b775Jvl8nnPnznHixAmKxWLlw57MGcz/8Fdvzw+LZCijYGguCoYVysxoa2tjy5Yt9Pf3s3XrVsyM119/nd7eXqampigUChw9epR3332X9vZ2pqenKZVKlWsmpqeno57B+917IQmGK12dKc1Dk48rWFtbG4ODg3R2dnLHHXfwiU98ggceeKAyAZlKpdi4cSOpVIpisUipVGJmZoapqak5ZxOqhw/zz1IsdFFV9VmK5AIr3cyluajHsEIlpyS7u7vp7e2lu7ubXC7H0NAQ2WyW06dPs3//fjZu3AhAR0dH5dLpiYkJ2tvbaW9vn/NBn391JRCFRfK1+pLuUqmk9QxNRsGwQiV/rZMPaLFYrAwPuru76enpYXR0lOnpaSYmJsjlcqRSKbLZbOVUZn9/P6Ojo1EPIfn+yf5CS6KTwEjOSGh+obloKLGCJYuXSqUSk5OTvPfee4yNjVEsFunq6uL222+no6ODnp4e+vr6KJVKtLW1VRY83XrrrZX96uHB/EnG+cOMJCSKxWLldKg0FwXDCtbT00NXVxfT09OMjIxw9OjRytmGQqHAwMAAk5OTtLe3Vz7E2WyW9evXc/HiRT784Q/T2dlJNpslk8lUhgZXCodE0mMoFApMT09rGNGEFAwrlJnR29tLe3s7AO+9917lpq2XL1+u3GchWQ2Z9AoymQxDQ0O0tbXR1dVFLperXKZ9pXCYP+GYSG4VJ81HwbBCmRnr1q2jtbWV2dlZxsfHyWazTExMzJkMTKfTdHR00NXVVVnGfOuttzIwMADAmjVr6OzspK+vr3LBVXJ36Ool0guFg85ENC8FwwplZrS3t1MqlTh16hQ9PT2cOXOG4eFhLl26xPnz55mamgKgu7ubdDpNW1sbra2t9Pf389GPfpR8Ps8HP/hBNmzYwJYtW9i0aRNdXV1ks9nKI3lN9TzEQhOS0lx0VmKFcnfOnTvH5OQkuVyO8+fPc/ToUS5dukRvby/Hjx/nwoULtLe3MzU1xeXLl8nlcpWzF+7OmjVrKkOJlpaWyiMJATOr9DzmTz4qFJqbgmGFKhaLHDp0iJtvvpn+/n6ee+45jhw5UjnzkM/nmZqaYnx8nJGREaamphgcHGTt2rW88MILpNNpurq6eOeddxgbGyOVSjExMVG5SWxyKjS550KydHqhNQ3SfBQMK1TSY/jNb35DZ2cnp06dYnx8nFQqxb59+xgfH2fdunVcvHiR4eFh8vk809PTTE1NsXHjRtyd4eFhjh8/zsTEBGbGzMwMMzMzlXUO1TdygbkTjwqF5qZgWKHcncuXLzM6Oko6nWZycpJCoYCZ8fvf/57x8XF6enqYmZlhcnKSYrHIhQsXGBwcZGhoiKmpKQ4fPszp06fn3A8yueQ66SUk6yQWugxbmpc1wi/RzJa/iFWk+q5NQGWeoK2tjba2NjKZDIVCgcnJycp8A8ztBcw/K5E8r7mGhnbQ3bfVcqB6DKtQ8sGdv8YguYgqWUKdXFh1pe+R3MRloTUM1Vdaajl087nq6Uozy5rZb83sVTM7ZGbfDO2bzexFMxs2syfNLBPa28L+cHh+qL4/glyP5ANb/Uj+v5X5fL7m6xuS11UHSfW9GPT/mGhOtaxjmAHuc/c/Au4E7jezu4FvAY+7+weA88Cj4fhHgfOh/fFwnDSB6rC4niHA/NdX/89zpblcNRi8bCLstoaHA/cBvwjte4CHwvb2sE94/jOmu3SsOtUBoV9/86lp5aOZpc3sFeA08CzwNjDu7skdPkeATWF7E3ACIDx/AVi7wPfcaWYHzOzAjf0I0qiSoYnmGJpPTZOP7l4E7jSzXuBXwIdu9I3dfTewG3RWYqXSfR6b1zVdK+Hu48DzwD1Ar5klwTIAjIbtUWAQIDzfA5xdlGpFZEnUclbiptBTwMxywGeBI5QD4uFw2A7g6bC9N+wTnt/vOpkt0lRqGUpsAPaYWZpykDzl7s+Y2WHg52b298DLwBPh+CeAfzWzYeAc8Egd6haROtLKR5HVo+aVj7ofg4hEFAwiElEwiEhEwSAiEQWDiEQUDCISUTCISETBICIRBYOIRBQMIhJRMIhIRMEgIhEFg4hEFAwiElEwiEhEwSAiEQWDiEQUDCISUTCISETBICIRBYOIRBQMIhJRMIhIRMEgIhEFg4hEFAwiElEwiEhEwSAiEQWDiEQUDCISUTCISETBICIRBYOIRGoOBjNLm9nLZvZM2N9sZi+a2bCZPWlmmdDeFvaHw/ND9SldROrlWnoMXwWOVO1/C3jc3T8AnAceDe2PAudD++PhOBFpIjUFg5kNAH8K/HPYN+A+4BfhkD3AQ2F7e9gnPP+ZcLyINIlaewzfAb4BlML+WmDc3WfD/giwKWxvAk4AhOcvhOPnMLOdZnbAzA5cZ+0iUidXDQYz+xxw2t0PLuYbu/tud9/m7tsW8/uKyI1rqeGYTwKfN7MHgSzQDXwX6DWzltArGABGw/GjwCAwYmYtQA9wdtErF5G6uWqPwd0fc/cBdx8CHgH2u/uXgOeBh8NhO4Cnw/besE94fr+7+6JWLSJ1dSPrGP4a+LqZDVOeQ3gitD8BrA3tXwd23ViJIrLUrBH+mJvZ8hchsvIdrHVOTysfRSSiYBCRiIJBRCIKBhGJKBhEJKJgEJGIgkFEIgoGEYkoGEQkomAQkYiCQUQiCgYRiSgYRCSiYBCRiIJBRCIKBhGJKBhEJKJgEJGIgkFEIgoGEYkoGEQkomAQkYiCQUQiCgYRiSgYRCSiYBCRiIJBRCIKBhGJKBhEJKJgEJGIgkFEIgoGEYkoGEQkomAQkUhNwWBmfzCz183sFTM7ENrWmNmzZvZW+NoX2s3Mvmdmw2b2mpndVc8fQEQW37X0GP7Y3e90921hfxewz923AvvCPsADwNbw2An8YLGKFZGlcSNDie3AnrC9B3ioqv3HXvYC0GtmG27gfURkidUaDA78t5kdNLOdoa3f3cfC9kmgP2xvAk5UvXYktM1hZjvN7EAyNBGRxtFS43GfcvdRM7sZeNbMjlY/6e5uZn4tb+zuu4HdANf6WhGpr5p6DO4+Gr6eBn4FfAw4lQwRwtfT4fBRYLDq5QOhTUSaxFWDwcw6zKwr2Qb+BHgD2AvsCIftAJ4O23uBL4ezE3cDF6qGHCLSBGoZSvQDvzKz5Pifuvt/mdlLwFNm9ihwHPhCOP4/gAeBYWAS+MqiVy0idWXuyz+8N7NLwLHlrqNG64D3lruIGjRLndA8tTZLnbBwrbe6+021vLjWycd6O1a1PqKhmdmBZqi1WeqE5qm1WeqEG69VS6JFJKJgEJFIowTD7uUu4Bo0S63NUic0T63NUifcYK0NMfkoIo2lUXoMItJAlj0YzOx+MzsWLtPedfVX1LWWH5nZaTN7o6qtIS8vN7NBM3vezA6b2SEz+2oj1mtmWTP7rZm9Gur8ZmjfbGYvhnqeNLNMaG8L+8Ph+aGlqLOq3rSZvWxmzzR4nfW9FYK7L9sDSANvA1uADPAqcNsy1nMvcBfwRlXbPwC7wvYu4Fth+0HgPwED7gZeXOJaNwB3he0u4E3gtkarN7xfZ9huBV4M7/8U8Eho/yHwZ2H7z4Efhu1HgCeX+L/r14GfAs+E/Uat8w/Aunlti/a7X7If5Ao/3D3Ar6v2HwMeW+aahuYFwzFgQ9jeQHnNBcA/AV9c6Lhlqvtp4LONXC/QDvwO+DjlxTct8/8dAL8G7gnbLeE4W6L6BijfW+Q+4JnwQWq4OsN7LhQMi/a7X+6hRE2XaC+zG7q8fCmEbuxHKP81brh6Q/f8FcoX2j1LuZc47u6zC9RSqTM8fwFYuxR1At8BvgGUwv7aBq0T6nArhGqNsvKxKbhf++Xl9WZmncC/A19z94vhmhagcep19yJwp5n1Ur4690PLXFLEzD4HnHb3g2b26eWupwaLfiuEasvdY2iGS7Qb9vJyM2ulHAo/cfdfhuaGrdfdx4HnKXfJe80s+cNUXUulzvB8D3B2Ccr7JPB5M/sD8HPKw4nvNmCdQP1vhbDcwfASsDXM/GYoT+LsXeaa5mvIy8ut3DV4Ajji7t9u1HrN7KbQU8DMcpTnQY5QDoiHr1BnUv/DwH4PA+N6cvfH3H3A3Yco/zvc7+5farQ6YYluhbBUkyXvM4nyIOUZ9beBv13mWn4GjAEFyuOwRymPG/cBbwHPAWvCsQZ8P9T9OrBtiWv9FOVx5mvAK+HxYKPVC9wBvBzqfAP4u9C+Bfgt5cvz/w1oC+3ZsD8cnt+yDP8OPs3/n5VouDpDTa+Gx6Hkc7OYv3utfBSRyHIPJUSkASkYRCSiYBCRiIJBRCIKBhGJKBhEJKJgEJGIgkFEIv8HQLl5G8NGbFMAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "t=10\n",
    "\n",
    "pred_img=model.predict(x_train[t:(t+1),...])\n",
    "\n",
    "print (pred_img.shape)\n",
    "\n",
    "plt.imshow(pred_img[0,:,:,0],cmap=plt.cm.gray)\n",
    "plt.show()\n",
    "\n",
    "# print (pred_img[0,:,:,0])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python2(conda)",
   "language": "python",
   "name": "dinghow"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}