image_retrieval.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "source": [
    "Import packages"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import keras\n",
    "from keras import layers\n",
    "import numpy as np\n",
    "import os\n",
    "import cv2\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "markdown",
   "source": [
    "Load data"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "outputs": [],
   "source": [
    "img_height = 512\n",
    "img_width = 512\n",
    "data_dir = \"data/Reference images\"\n",
    "batch_size = 20\n",
    "\n",
    "images = []\n",
    "for filename in os.listdir(data_dir):\n",
    "\n",
    "    img = cv2.imread(os.path.join(data_dir,filename),cv2.IMREAD_GRAYSCALE)\n",
    "    if img is not None:\n",
    "        images.append(img)\n"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "Visualize the data"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "outputs": [
    {
     "data": {
      "text/plain": "<Figure size 640x480 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(cv2.cvtColor(images[0], cv2.COLOR_BGR2RGB))\n",
    "plt.show()"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "Preprocess images"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "outputs": [
    {
     "data": {
      "text/plain": "array([[[[1.],\n         [1.],\n         [1.],\n         ...,\n         [1.],\n         [1.],\n         [1.]],\n\n        [[1.],\n         [1.],\n         [1.],\n         ...,\n         [1.],\n         [1.],\n         [1.]],\n\n        [[1.],\n         [1.],\n         [1.],\n         ...,\n         [1.],\n         [1.],\n         [1.]],\n\n        ...,\n\n        [[1.],\n         [1.],\n         [1.],\n         ...,\n         [1.],\n         [1.],\n         [1.]],\n\n        [[1.],\n         [1.],\n         [1.],\n         ...,\n         [1.],\n         [1.],\n         [1.]],\n\n        [[1.],\n         [1.],\n         [1.],\n         ...,\n         [1.],\n         [1.],\n         [1.]]],\n\n\n       [[[1.],\n         [1.],\n         [1.],\n         ...,\n         [1.],\n         [1.],\n         [1.]],\n\n        [[1.],\n         [1.],\n         [1.],\n         ...,\n         [1.],\n         [1.],\n         [1.]],\n\n        [[1.],\n         [1.],\n         [1.],\n         ...,\n         [1.],\n         [1.],\n         [1.]],\n\n        ...,\n\n        [[1.],\n         [1.],\n         [1.],\n         ...,\n         [1.],\n         [1.],\n         [1.]],\n\n        [[1.],\n         [1.],\n         [1.],\n         ...,\n         [1.],\n         [1.],\n         [1.]],\n\n        [[1.],\n         [1.],\n         [1.],\n         ...,\n         [1.],\n         [1.],\n         [1.]]],\n\n\n       [[[1.],\n         [1.],\n         [1.],\n         ...,\n         [1.],\n         [1.],\n         [1.]],\n\n        [[1.],\n         [1.],\n         [1.],\n         ...,\n         [1.],\n         [1.],\n         [1.]],\n\n        [[1.],\n         [1.],\n         [1.],\n         ...,\n         [1.],\n         [1.],\n         [1.]],\n\n        ...,\n\n        [[1.],\n         [1.],\n         [1.],\n         ...,\n         [1.],\n         [1.],\n         [1.]],\n\n        [[1.],\n         [1.],\n         [1.],\n         ...,\n         [1.],\n         [1.],\n         [1.]],\n\n        [[1.],\n         [1.],\n         [1.],\n         ...,\n         [1.],\n         [1.],\n         [1.]]]])"
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_images = images[:int(len(images)*0.9)]\n",
    "test_images = images[int(len(images)*0.9):]\n",
    "\n",
    "\n",
    "train_images = np.array(train_images) / 255.\n",
    "test_images = np.array(test_images) / 255.\n",
    "\n",
    "train_images = np.reshape(train_images,(len(train_images), img_height, img_width, 1))\n",
    "test_images.reshape(len(test_images), img_height, img_width, 1)\n",
    "\n"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "Define the model"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"model\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " input_1 (InputLayer)        [(None, 512, 512, 1)]     0         \n",
      "                                                                 \n",
      " conv2d (Conv2D)             (None, 512, 512, 256)     2560      \n",
      "                                                                 \n",
      " max_pooling2d (MaxPooling2D  (None, 256, 256, 256)    0         \n",
      " )                                                               \n",
      "                                                                 \n",
      " conv2d_1 (Conv2D)           (None, 256, 256, 128)     295040    \n",
      "                                                                 \n",
      " max_pooling2d_1 (MaxPooling  (None, 128, 128, 128)    0         \n",
      " 2D)                                                             \n",
      "                                                                 \n",
      " conv2d_2 (Conv2D)           (None, 128, 128, 64)      73792     \n",
      "                                                                 \n",
      " max_pooling2d_2 (MaxPooling  (None, 64, 64, 64)       0         \n",
      " 2D)                                                             \n",
      "                                                                 \n",
      " conv2d_3 (Conv2D)           (None, 64, 64, 32)        18464     \n",
      "                                                                 \n",
      " max_pooling2d_3 (MaxPooling  (None, 32, 32, 32)       0         \n",
      " 2D)                                                             \n",
      "                                                                 \n",
      " conv2d_4 (Conv2D)           (None, 32, 32, 32)        9248      \n",
      "                                                                 \n",
      " up_sampling2d (UpSampling2D  (None, 64, 64, 32)       0         \n",
      " )                                                               \n",
      "                                                                 \n",
      " conv2d_5 (Conv2D)           (None, 64, 64, 64)        18496     \n",
      "                                                                 \n",
      " up_sampling2d_1 (UpSampling  (None, 128, 128, 64)     0         \n",
      " 2D)                                                             \n",
      "                                                                 \n",
      " conv2d_6 (Conv2D)           (None, 128, 128, 128)     73856     \n",
      "                                                                 \n",
      " up_sampling2d_2 (UpSampling  (None, 256, 256, 128)    0         \n",
      " 2D)                                                             \n",
      "                                                                 \n",
      " conv2d_7 (Conv2D)           (None, 256, 256, 256)     295168    \n",
      "                                                                 \n",
      " up_sampling2d_3 (UpSampling  (None, 512, 512, 256)    0         \n",
      " 2D)                                                             \n",
      "                                                                 \n",
      " conv2d_8 (Conv2D)           (None, 512, 512, 1)       2305      \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 788,929\n",
      "Trainable params: 788,929\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "input_img = keras.Input(shape=(512, 512, 1))\n",
    "# ----------------------- ENCODER —---------------\n",
    "x = layers.Conv2D(256, (3, 3), activation='relu', padding='same')(input_img)\n",
    "x = layers.MaxPooling2D((2, 2), padding='same')(x)\n",
    "x = layers.Conv2D(128, (3, 3), activation='relu', padding='same')(x)\n",
    "x = layers.MaxPooling2D((2, 2), padding='same')(x)\n",
    "x = layers.Conv2D(64, (3, 3), activation='relu', padding='same')(x)\n",
    "x = layers.MaxPooling2D((2, 2), padding='same')(x)\n",
    "x = layers.Conv2D(32, (3, 3), activation='relu', padding='same')(x)\n",
    "encoded = layers.MaxPooling2D((2, 2), padding='same')(x)\n",
    "\n",
    "# ----------------------- DECODER —---------------\n",
    "x = layers.Conv2D(32, (3, 3), activation='relu', padding='same')(encoded)\n",
    "x = layers.UpSampling2D((2, 2))(x)\n",
    "x = layers.Conv2D(64, (3, 3), activation='relu', padding='same')(x)\n",
    "x = layers.UpSampling2D((2, 2))(x)\n",
    "x = layers.Conv2D(128, (3, 3), activation='relu', padding='same')(x)\n",
    "x = layers.UpSampling2D((2, 2))(x)\n",
    "x = layers.Conv2D(256, (3, 3), activation='relu', padding='same')(x)\n",
    "x = layers.UpSampling2D((2, 2))(x)\n",
    "decoded = layers.Conv2D(1, (3, 3), activation='sigmoid', padding='same')(x)\n",
    "\n",
    "autoencoder = keras.Model(input_img, decoded)\n",
    "autoencoder.summary()"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "Compile the model"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "outputs": [],
   "source": [
    "autoencoder.compile(optimizer='adam', loss='binary_crossentropy')"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "Fit the model"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "autoencoder.fit(train_images, train_images, epochs=10)"
   ],
   "metadata": {
    "collapsed": false
   }
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}