Gaussian Label Noise.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "rolled-subscription",
   "metadata": {},
   "outputs": [],
   "source": [
    "import warnings\n",
    "from osgeo import gdal\n",
    "from ipywidgets import IntProgress\n",
    "import glob\n",
    "import random\n",
    "import cv2\n",
    "import numpy as np\n",
    "import tensorflow as tf\n",
    "import rasterio\n",
    "from rasterio.plot import show\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "import os\n",
    "from sklearn.model_selection import train_test_split\n",
    "from PIL import Image\n",
    "from tensorflow.keras.layers import Conv2D, Activation, BatchNormalization\n",
    "from tensorflow.keras.layers import UpSampling2D, Input, Concatenate\n",
    "from tensorflow.keras.models import Model\n",
    "from tensorflow.keras.applications import MobileNetV2\n",
    "from tensorflow.keras.callbacks import EarlyStopping, ReduceLROnPlateau\n",
    "from tensorflow.keras.metrics import Recall, Precision\n",
    "import tensorflow.keras.backend as K\n",
    "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
    "from tensorflow.keras import layers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "naked-surname",
   "metadata": {},
   "outputs": [],
   "source": [
    "seed = 42\n",
    "lr = 1e-4"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "square-fever",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "colonial-discussion",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of image samples: 12762\n",
      "Number of mask samples: 12762\n"
     ]
    }
   ],
   "source": [
    "import os\n",
    "\n",
    "input_dir = \"./data/Data/Dec19_image/\"\n",
    "target_dir = \"./data/Data/Dec19_mask/\"\n",
    "img_size = (256, 256)\n",
    "num_classes = 5\n",
    "batch_size = 32\n",
    "\n",
    "input_img_paths = sorted(\n",
    "    [\n",
    "        os.path.join(input_dir, fname)\n",
    "        for fname in os.listdir(input_dir)\n",
    "        if fname.endswith(\".tif\")\n",
    "    ]\n",
    ")\n",
    "\n",
    "target_img_paths = sorted(\n",
    "    [\n",
    "        os.path.join(target_dir, fname)\n",
    "        for fname in os.listdir(target_dir)\n",
    "        if fname.endswith(\".tif\")\n",
    "    ]\n",
    ")\n",
    "\n",
    "n = 1500\n",
    "input_img_paths = input_img_paths[n:]\n",
    "target_img_paths = target_img_paths[n:]\n",
    "input_img_paths = input_img_paths[:-n]\n",
    "target_img_paths = target_img_paths[:-n]\n",
    "\n",
    "print(\"Number of image samples:\", len(input_img_paths))\n",
    "\n",
    "print(\"Number of mask samples:\", len(target_img_paths))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "drawn-laser",
   "metadata": {},
   "outputs": [],
   "source": [
    "def read_original_mask(path):\n",
    "    img = gdal.Open(path)\n",
    "    x = img.ReadAsArray()\n",
    "    x = np.array(x)\n",
    "    x = np.array(Image.fromarray(x.astype(np.uint8)).resize((256, 256)))\n",
    "    x[np.isnan(x)] = 0\n",
    "    x = np.expand_dims(x, -1)\n",
    "    #x = to_categorical(x,num_classes=5)\n",
    "    #x = np.asarray(x, dtype=np.float32)\n",
    "    print(x.shape, \"read mask\")\n",
    "    return x\n",
    "\n",
    "\n",
    "# land - green\n",
    "# high - red\n",
    "# moderate - yellow\n",
    "# low - black\n",
    "# water - blue\n",
    "\n",
    "def make_colour_mask(segment_mask):\n",
    "    colour_table = np.zeros((256, 1, 3), dtype=np.uint8)\n",
    "    colour_table[0] = [0,128,0] \n",
    "    colour_table[1] = [255,0,0]\n",
    "    colour_table[2] = [255,255,0]\n",
    "    colour_table[3] = [0,0,0]\n",
    "    colour_table[4] = [0,0,255]\n",
    "\n",
    "    colour_mask = cv2.applyColorMap(segment_mask, colour_table)\n",
    "\n",
    "    return colour_mask"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "studied-thong",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(256, 256, 1) read mask\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x234b9b2afd0>"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "ind = 1512\n",
    "fp = input_img_paths[ind]\n",
    "img = rasterio.open(fp)\n",
    "show(img)\n",
    "\n",
    "\n",
    "\n",
    "file = target_img_paths[ind]\n",
    "im = read_original_mask(file)\n",
    "img = make_colour_mask(im)\n",
    "plt.imshow(img)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "medieval-syndicate",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import LabelEncoder\n",
    "from keras.utils import to_categorical\n",
    "\n",
    "# Normalize bands into 0.0 - 1.0 scale\n",
    "def normalize(array):\n",
    "    diff = array.max() - array.min()\n",
    "    return (array - array.min()) / diff\n",
    "\n",
    "def read_image(path):\n",
    "    raster = rasterio.open(path)\n",
    "    band_2 = raster.read(1)\n",
    "    band_3 = raster.read(2)\n",
    "    band_4 = raster.read(3)\n",
    "    \n",
    "    norm_2 = normalize(band_2)\n",
    "    norm_3 = normalize(band_3)\n",
    "    norm_4 = normalize(band_4)\n",
    "    \n",
    "    x = np.dstack((norm_4, norm_3, norm_2))\n",
    "    image = np.array(x)\n",
    "    image = np.array(Image.fromarray((x * 255).astype(np.uint8)).resize((256, 256)).convert('RGB'))\n",
    "    image[np.isnan(image)] = 0\n",
    "    image = np.asarray(image, dtype=np.float32)/255.0\n",
    "    #print(\"image\")\n",
    "    #print(image.shape)\n",
    "    #print(image)\n",
    "    return image\n",
    "\n",
    "def read_mask(path):\n",
    "    img = gdal.Open(path)\n",
    "    x = img.ReadAsArray()\n",
    "    x = np.array(x)\n",
    "    x = np.array(Image.fromarray(x.astype(np.uint8)).resize((256, 256)))\n",
    "    x[np.isnan(x)] = 0\n",
    "    x = np.expand_dims(x, -1)\n",
    "    x = to_categorical(x,num_classes=5)\n",
    "    x = np.asarray(x, dtype=np.float32)\n",
    "    # print(x.shape, \"read mask\")\n",
    "    return x\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "violent-strain",
   "metadata": {},
   "outputs": [],
   "source": [
    "from random import randrange\n",
    "\n",
    "def readnoise(path):\n",
    "    img = gdal.Open(path)\n",
    "    x = img.ReadAsArray()\n",
    "    x = np.array(x)\n",
    "    x = np.array(Image.fromarray(x.astype(np.uint8)).resize((256, 256)))\n",
    "    x[np.isnan(x)] = 0\n",
    "    x = np.expand_dims(x, -1)\n",
    "    x = np.asarray(x, dtype=np.float32)\n",
    "    \n",
    "    for i in range(6550): \n",
    "        pix = randrange(5)\n",
    "        pix = np.float32(pix)\n",
    "        rand1 = randrange(256)\n",
    "        rand2 = randrange(256) \n",
    "        x[rand1][rand2] = pix\n",
    "    \n",
    "    x = to_categorical(x,num_classes=5)\n",
    "    #print(x, \" \", x.shape)\n",
    "    return x\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "together-cabin",
   "metadata": {},
   "outputs": [],
   "source": [
    "im = readnoise(target_img_paths[120])\n",
    "#im = noise1per(im)\n",
    "#print(\"Min and Max values of the images : \", im.max(), im.min())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "sacred-hearts",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Shape of the iamges  (256, 256, 3)\n",
      "Min and Max values of the images :  1.0 0.0\n"
     ]
    }
   ],
   "source": [
    "im2 = read_image(input_img_paths[310])\n",
    "\n",
    "print(\"Shape of the iamges \", im2.shape)\n",
    "print(\"Min and Max values of the images : \", im2.max(), im2.min())\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "unexpected-conspiracy",
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow import keras\n",
    "import numpy as np\n",
    "from tensorflow.keras.preprocessing.image import load_img\n",
    "\n",
    "from sklearn.preprocessing import LabelEncoder\n",
    "from keras.utils import to_categorical\n",
    "\n",
    "\n",
    "class SATDATA(keras.utils.Sequence):\n",
    "    \"\"\"Helper to iterate over the data (as Numpy arrays).\"\"\"\n",
    "\n",
    "    def __init__(self, batch_size, img_size, input_img_paths, target_img_paths):\n",
    "        self.batch_size = batch_size\n",
    "        self.img_size = img_size\n",
    "        self.input_img_paths = input_img_paths\n",
    "        self.target_img_paths = target_img_paths\n",
    "\n",
    "    def __len__(self):\n",
    "        return len(self.target_img_paths) // self.batch_size\n",
    "\n",
    "    def __getitem__(self, idx):\n",
    "        \"\"\"Returns tuple (input, target) correspond to batch #idx.\"\"\"\n",
    "        i = idx * self.batch_size\n",
    "        batch_input_img_paths = self.input_img_paths[i : i + self.batch_size]\n",
    "        batch_target_img_paths = self.target_img_paths[i : i + self.batch_size]\n",
    "        x = np.zeros((self.batch_size,) + self.img_size + (3,), dtype=\"float32\")\n",
    "        \n",
    "        for j, path in enumerate(batch_input_img_paths):\n",
    "            #print(path)\n",
    "            img = read_image(path)\n",
    "            x[j] = img\n",
    "            x[j] = np.stack(img)\n",
    "            \n",
    "        y = np.zeros((self.batch_size,) + self.img_size + (5,), dtype=\"float32\")\n",
    "        for j, path in enumerate(batch_target_img_paths):\n",
    "            y[j] = readnoise(path)\n",
    "         \n",
    "        #print(x.shape, \"  \", y.shape)\n",
    "        return x, y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "premier-flight",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "beautiful-viking",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "fifteen-throw",
   "metadata": {},
   "outputs": [],
   "source": [
    "import random\n",
    "\n",
    "\n",
    "# Split our img paths into a training and a validation set\n",
    "val_samples = 3000\n",
    "random.Random(1337).shuffle(input_img_paths)\n",
    "random.Random(1337).shuffle(target_img_paths)\n",
    "\n",
    "train_input_img_paths = input_img_paths[:-val_samples]\n",
    "train_target_img_paths = target_img_paths[:-val_samples]\n",
    "\n",
    "# print(train_input_img_paths[100])\n",
    "# print(train_target_img_paths[100])\n",
    "\n",
    "val_input_img_paths = input_img_paths[-val_samples:]\n",
    "val_target_img_paths = target_img_paths[-val_samples:]\n",
    "\n",
    "# Instantiate data Sequences for each split\n",
    "train_gen = SATDATA(batch_size, img_size, train_input_img_paths, train_target_img_paths)\n",
    "val_gen = SATDATA(batch_size, img_size, val_input_img_paths, val_target_img_paths)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "canadian-projector",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "skilled-person",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "93\n",
      "305\n"
     ]
    }
   ],
   "source": [
    "print(val_gen.__len__())\n",
    "print(train_gen.__len__())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "innocent-miami",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-----------------------------------------\n",
      "keras-unet init: TF version is >= 2.0.0 - using `tf.keras` instead of `Keras`\n",
      "-----------------------------------------\n",
      "Model: \"model\"\n",
      "__________________________________________________________________________________________________\n",
      "Layer (type)                    Output Shape         Param #     Connected to                     \n",
      "==================================================================================================\n",
      "input_1 (InputLayer)            [(None, 256, 256, 3) 0                                            \n",
      "__________________________________________________________________________________________________\n",
      "conv2d (Conv2D)                 (None, 256, 256, 64) 1792        input_1[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization (BatchNorma (None, 256, 256, 64) 256         conv2d[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_1 (Conv2D)               (None, 256, 256, 64) 36928       batch_normalization[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_1 (BatchNor (None, 256, 256, 64) 256         conv2d_1[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_2 (Conv2D)               (None, 256, 256, 64) 36928       batch_normalization_1[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling2d (MaxPooling2D)    (None, 128, 128, 64) 0           conv2d_2[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_2 (BatchNor (None, 128, 128, 64) 256         max_pooling2d[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_3 (Conv2D)               (None, 128, 128, 64) 36928       batch_normalization_2[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_3 (BatchNor (None, 128, 128, 64) 256         conv2d_3[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_4 (Conv2D)               (None, 128, 128, 64) 36928       batch_normalization_3[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_4 (BatchNor (None, 128, 128, 64) 256         conv2d_4[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_5 (Conv2D)               (None, 128, 128, 64) 36928       batch_normalization_4[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling2d_1 (MaxPooling2D)  (None, 64, 64, 64)   0           conv2d_5[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_5 (BatchNor (None, 64, 64, 64)   256         max_pooling2d_1[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_6 (Conv2D)               (None, 64, 64, 64)   36928       batch_normalization_5[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_6 (BatchNor (None, 64, 64, 64)   256         conv2d_6[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_7 (Conv2D)               (None, 64, 64, 64)   36928       batch_normalization_6[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_7 (BatchNor (None, 64, 64, 64)   256         conv2d_7[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_8 (Conv2D)               (None, 64, 64, 64)   36928       batch_normalization_7[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling2d_2 (MaxPooling2D)  (None, 32, 32, 64)   0           conv2d_8[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_8 (BatchNor (None, 32, 32, 64)   256         max_pooling2d_2[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_9 (Conv2D)               (None, 32, 32, 64)   36928       batch_normalization_8[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_9 (BatchNor (None, 32, 32, 64)   256         conv2d_9[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_10 (Conv2D)              (None, 32, 32, 64)   36928       batch_normalization_9[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_10 (BatchNo (None, 32, 32, 64)   256         conv2d_10[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_11 (Conv2D)              (None, 32, 32, 64)   36928       batch_normalization_10[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling2d_3 (MaxPooling2D)  (None, 16, 16, 64)   0           conv2d_11[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_11 (BatchNo (None, 16, 16, 64)   256         max_pooling2d_3[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_12 (Conv2D)              (None, 16, 16, 64)   36928       batch_normalization_11[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_12 (BatchNo (None, 16, 16, 64)   256         conv2d_12[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_13 (Conv2D)              (None, 16, 16, 64)   36928       batch_normalization_12[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_13 (BatchNo (None, 16, 16, 64)   256         conv2d_13[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_14 (Conv2D)              (None, 16, 16, 64)   36928       batch_normalization_13[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling2d_4 (MaxPooling2D)  (None, 8, 8, 64)     0           conv2d_14[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_14 (BatchNo (None, 8, 8, 64)     256         max_pooling2d_4[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_15 (Conv2D)              (None, 8, 8, 64)     36928       batch_normalization_14[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_15 (BatchNo (None, 8, 8, 64)     256         conv2d_15[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_16 (Conv2D)              (None, 8, 8, 64)     36928       batch_normalization_15[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_16 (BatchNo (None, 8, 8, 64)     256         conv2d_16[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_transpose (Conv2DTranspo (None, 16, 16, 64)   36928       batch_normalization_16[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "concatenate (Concatenate)       (None, 16, 16, 128)  0           conv2d_transpose[0][0]           \n",
      "                                                                 conv2d_13[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_17 (BatchNo (None, 16, 16, 128)  512         concatenate[0][0]                \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_17 (Conv2D)              (None, 16, 16, 96)   110688      batch_normalization_17[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_18 (BatchNo (None, 16, 16, 96)   384         conv2d_17[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_18 (Conv2D)              (None, 16, 16, 64)   55360       batch_normalization_18[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_19 (BatchNo (None, 16, 16, 64)   256         conv2d_18[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_transpose_1 (Conv2DTrans (None, 32, 32, 64)   36928       batch_normalization_19[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_1 (Concatenate)     (None, 32, 32, 128)  0           conv2d_transpose_1[0][0]         \n",
      "                                                                 conv2d_10[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_20 (BatchNo (None, 32, 32, 128)  512         concatenate_1[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_19 (Conv2D)              (None, 32, 32, 96)   110688      batch_normalization_20[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_21 (BatchNo (None, 32, 32, 96)   384         conv2d_19[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_20 (Conv2D)              (None, 32, 32, 64)   55360       batch_normalization_21[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_22 (BatchNo (None, 32, 32, 64)   256         conv2d_20[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_transpose_2 (Conv2DTrans (None, 64, 64, 64)   36928       batch_normalization_22[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_2 (Concatenate)     (None, 64, 64, 128)  0           conv2d_transpose_2[0][0]         \n",
      "                                                                 conv2d_7[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_23 (BatchNo (None, 64, 64, 128)  512         concatenate_2[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_21 (Conv2D)              (None, 64, 64, 96)   110688      batch_normalization_23[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_24 (BatchNo (None, 64, 64, 96)   384         conv2d_21[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_22 (Conv2D)              (None, 64, 64, 64)   55360       batch_normalization_24[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_25 (BatchNo (None, 64, 64, 64)   256         conv2d_22[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_transpose_3 (Conv2DTrans (None, 128, 128, 64) 36928       batch_normalization_25[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_3 (Concatenate)     (None, 128, 128, 128 0           conv2d_transpose_3[0][0]         \n",
      "                                                                 conv2d_4[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_26 (BatchNo (None, 128, 128, 128 512         concatenate_3[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_23 (Conv2D)              (None, 128, 128, 96) 110688      batch_normalization_26[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_27 (BatchNo (None, 128, 128, 96) 384         conv2d_23[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_24 (Conv2D)              (None, 128, 128, 64) 55360       batch_normalization_27[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_28 (BatchNo (None, 128, 128, 64) 256         conv2d_24[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_transpose_4 (Conv2DTrans (None, 256, 256, 64) 36928       batch_normalization_28[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_4 (Concatenate)     (None, 256, 256, 128 0           conv2d_transpose_4[0][0]         \n",
      "                                                                 conv2d_1[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_29 (BatchNo (None, 256, 256, 128 512         concatenate_4[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_25 (Conv2D)              (None, 256, 256, 96) 110688      batch_normalization_29[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_30 (BatchNo (None, 256, 256, 96) 384         conv2d_25[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_26 (Conv2D)              (None, 256, 256, 64) 55360       batch_normalization_30[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_27 (Conv2D)              (None, 256, 256, 5)  325         conv2d_26[0][0]                  \n",
      "==================================================================================================\n",
      "Total params: 1,617,701\n",
      "Trainable params: 1,612,773\n",
      "Non-trainable params: 4,928\n",
      "__________________________________________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "# https://cdn-sv1.deepsense.ai/wp-content/uploads/2017/04/architecture_details.png\n",
    "# https://deepsense.ai/deep-learning-for-satellite-imagery-via-image-segmentation/\n",
    "\n",
    "from keras_unet import TF\n",
    "if TF:\n",
    "    from tensorflow.keras.models import Model\n",
    "    from tensorflow.keras.layers import ( \n",
    "        BatchNormalization, Conv2D, Conv2DTranspose,\n",
    "        MaxPooling2D, UpSampling2D, Input,\n",
    "        concatenate\n",
    "    )\n",
    "else:    \n",
    "    from keras.models import Model\n",
    "    from keras.layers import ( \n",
    "        BatchNormalization, Conv2D, Conv2DTranspose,\n",
    "        MaxPooling2D, UpSampling2D, Input,\n",
    "        concatenate\n",
    "    )\n",
    "\n",
    "\n",
    "def bn_conv_relu(input, filters, bachnorm_momentum, **conv2d_args):\n",
    "    x = BatchNormalization(momentum=bachnorm_momentum)(input)\n",
    "    x = Conv2D(filters, **conv2d_args)(x)\n",
    "    return x\n",
    "\n",
    "def bn_upconv_relu(input, filters, bachnorm_momentum, **conv2d_trans_args):\n",
    "    x = BatchNormalization(momentum=bachnorm_momentum)(input)\n",
    "    x = Conv2DTranspose(filters, **conv2d_trans_args)(x)\n",
    "    return x\n",
    "\n",
    "def satellite_unet(\n",
    "    input_shape,\n",
    "    num_classes=5,\n",
    "    output_activation='softmax',\n",
    "    num_layers=4):\n",
    "\n",
    "    inputs = Input(input_shape)   \n",
    "    \n",
    "    filters = 64\n",
    "    upconv_filters = 96\n",
    "\n",
    "    kernel_size = (3,3)\n",
    "    activation = 'relu'\n",
    "    strides = (1,1)\n",
    "    padding = 'same'\n",
    "    kernel_initializer = 'he_normal'\n",
    "\n",
    "    conv2d_args = {\n",
    "        'kernel_size':kernel_size,\n",
    "        'activation':activation, \n",
    "        'strides':strides,\n",
    "        'padding':padding,\n",
    "        'kernel_initializer':kernel_initializer\n",
    "        }\n",
    "\n",
    "    conv2d_trans_args = {\n",
    "        'kernel_size':kernel_size,\n",
    "        'activation':activation, \n",
    "        'strides':(2,2),\n",
    "        'padding':padding,\n",
    "        'output_padding':(1,1)\n",
    "        }\n",
    "\n",
    "    bachnorm_momentum = 0.01\n",
    "\n",
    "    pool_size = (2,2)\n",
    "    pool_strides = (2,2)\n",
    "    pool_padding = 'valid'\n",
    "\n",
    "    maxpool2d_args = {\n",
    "        'pool_size':pool_size,\n",
    "        'strides':pool_strides,\n",
    "        'padding':pool_padding,\n",
    "        }\n",
    "    \n",
    "    x = Conv2D(filters, **conv2d_args)(inputs)\n",
    "    c1 = bn_conv_relu(x, filters, bachnorm_momentum, **conv2d_args)    \n",
    "    x = bn_conv_relu(c1, filters, bachnorm_momentum, **conv2d_args)\n",
    "    x = MaxPooling2D(**maxpool2d_args)(x)\n",
    "\n",
    "    down_layers = []\n",
    "\n",
    "    for l in range(num_layers):\n",
    "        x = bn_conv_relu(x, filters, bachnorm_momentum, **conv2d_args)\n",
    "        x = bn_conv_relu(x, filters, bachnorm_momentum, **conv2d_args)\n",
    "        down_layers.append(x)\n",
    "        x = bn_conv_relu(x, filters, bachnorm_momentum, **conv2d_args)\n",
    "        x = MaxPooling2D(**maxpool2d_args)(x)\n",
    "\n",
    "    x = bn_conv_relu(x, filters, bachnorm_momentum, **conv2d_args)\n",
    "    x = bn_conv_relu(x, filters, bachnorm_momentum, **conv2d_args)\n",
    "    x = bn_upconv_relu(x, filters, bachnorm_momentum, **conv2d_trans_args)\n",
    "\n",
    "    for conv in reversed(down_layers):        \n",
    "        x = concatenate([x, conv])  \n",
    "        x = bn_conv_relu(x, upconv_filters, bachnorm_momentum, **conv2d_args)\n",
    "        x = bn_conv_relu(x, filters, bachnorm_momentum, **conv2d_args)\n",
    "        x = bn_upconv_relu(x, filters, bachnorm_momentum, **conv2d_trans_args)\n",
    "\n",
    "    x = concatenate([x, c1])\n",
    "    x = bn_conv_relu(x, upconv_filters, bachnorm_momentum, **conv2d_args)\n",
    "    x = bn_conv_relu(x, filters, bachnorm_momentum, **conv2d_args)\n",
    "           \n",
    "    outputs = Conv2D(5, kernel_size=(1,1), strides=(1,1), activation=output_activation, padding='valid') (x)   \n",
    "    \n",
    "    #print(\"output data\")\n",
    "    #print(outputs.shape)\n",
    "    \n",
    "    model = Model(inputs=[inputs], outputs=[outputs])\n",
    "    return model\n",
    "\n",
    "\n",
    "\n",
    "# Free up RAM in case the model definition cells were run multiple times\n",
    "keras.backend.clear_session()\n",
    "\n",
    "# Build model\n",
    "model = satellite_unet(input_shape=(256, 256, 3))\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "handled-strength",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "growing-grocery",
   "metadata": {},
   "outputs": [],
   "source": [
    "def dice_coef(y_true, y_pred, smooth=1):\n",
    "    \"\"\"\n",
    "    Dice = (2*|X & Y|)/ (|X|+ |Y|)\n",
    "         =  2*sum(|A*B|)/(sum(A^2)+sum(B^2))\n",
    "    ref: https://arxiv.org/pdf/1606.04797v1.pdf\n",
    "    \"\"\"\n",
    "    intersection = K.sum(K.abs(y_true * y_pred), axis=-1)\n",
    "    return (2. * intersection + smooth) / (K.sum(K.square(y_true),-1) + K.sum(K.square(y_pred),-1) + smooth)\n",
    "\n",
    "def dice_coef_loss(y_true, y_pred):\n",
    "    return 1-dice_coef(y_true, y_pred)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "elect-suggestion",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/30\n",
      "9/9 [==============================] - 1484s 179s/step - loss: 1.9831 - dice_coef: 0.6197 - accuracy: 0.2773 - val_loss: 1.5322 - val_dice_coef: 0.6941 - val_accuracy: 0.4206\n",
      "\n",
      "Epoch 00001: val_loss improved from inf to 1.53222, saving model to Dec19_noise10.h5\n",
      "Epoch 2/30\n",
      "9/9 [==============================] - 1520s 183s/step - loss: 1.4383 - dice_coef: 0.7092 - accuracy: 0.4537 - val_loss: 1.3872 - val_dice_coef: 0.7402 - val_accuracy: 0.5280\n",
      "\n",
      "Epoch 00002: val_loss improved from 1.53222 to 1.38718, saving model to Dec19_noise10.h5\n",
      "Epoch 3/30\n",
      "9/9 [==============================] - 1516s 182s/step - loss: 1.3801 - dice_coef: 0.7290 - accuracy: 0.5035 - val_loss: 1.2796 - val_dice_coef: 0.7542 - val_accuracy: 0.5504\n",
      "\n",
      "Epoch 00003: val_loss improved from 1.38718 to 1.27961, saving model to Dec19_noise10.h5\n",
      "Epoch 4/30\n",
      "9/9 [==============================] - 1515s 182s/step - loss: 1.3431 - dice_coef: 0.7361 - accuracy: 0.5135 - val_loss: 1.2265 - val_dice_coef: 0.7611 - val_accuracy: 0.5662\n",
      "\n",
      "Epoch 00004: val_loss improved from 1.27961 to 1.22654, saving model to Dec19_noise10.h5\n",
      "Epoch 5/30\n",
      "9/9 [==============================] - 1516s 182s/step - loss: 1.1990 - dice_coef: 0.7546 - accuracy: 0.5417 - val_loss: 1.2398 - val_dice_coef: 0.7513 - val_accuracy: 0.5423\n",
      "\n",
      "Epoch 00005: val_loss did not improve from 1.22654\n",
      "Epoch 6/30\n",
      "9/9 [==============================] - 1519s 183s/step - loss: 1.2221 - dice_coef: 0.7510 - accuracy: 0.5486 - val_loss: 1.1346 - val_dice_coef: 0.7730 - val_accuracy: 0.6059\n",
      "\n",
      "Epoch 00006: val_loss improved from 1.22654 to 1.13463, saving model to Dec19_noise10.h5\n",
      "Epoch 7/30\n",
      "9/9 [==============================] - 1520s 183s/step - loss: 1.1291 - dice_coef: 0.7736 - accuracy: 0.6079 - val_loss: 1.6947 - val_dice_coef: 0.7442 - val_accuracy: 0.5438\n",
      "\n",
      "Epoch 00007: val_loss did not improve from 1.13463\n",
      "Epoch 8/30\n",
      "9/9 [==============================] - 1517s 182s/step - loss: 1.1527 - dice_coef: 0.7682 - accuracy: 0.5943 - val_loss: 1.0637 - val_dice_coef: 0.7876 - val_accuracy: 0.6313\n",
      "\n",
      "Epoch 00008: val_loss improved from 1.13463 to 1.06368, saving model to Dec19_noise10.h5\n",
      "Epoch 9/30\n",
      "9/9 [==============================] - 1523s 183s/step - loss: 1.0659 - dice_coef: 0.7847 - accuracy: 0.6279 - val_loss: 1.0331 - val_dice_coef: 0.7913 - val_accuracy: 0.6390\n",
      "\n",
      "Epoch 00009: val_loss improved from 1.06368 to 1.03308, saving model to Dec19_noise10.h5\n",
      "Epoch 10/30\n",
      "9/9 [==============================] - 1538s 185s/step - loss: 1.0372 - dice_coef: 0.7915 - accuracy: 0.6452 - val_loss: 1.0661 - val_dice_coef: 0.7854 - val_accuracy: 0.6232\n",
      "\n",
      "Epoch 00010: val_loss did not improve from 1.03308\n",
      "Epoch 11/30\n",
      "9/9 [==============================] - 1529s 184s/step - loss: 1.0129 - dice_coef: 0.7946 - accuracy: 0.6414 - val_loss: 1.0225 - val_dice_coef: 0.7955 - val_accuracy: 0.6429\n",
      "\n",
      "Epoch 00011: val_loss improved from 1.03308 to 1.02250, saving model to Dec19_noise10.h5\n",
      "Epoch 12/30\n",
      "9/9 [==============================] - 1529s 184s/step - loss: 1.0943 - dice_coef: 0.7782 - accuracy: 0.6113 - val_loss: 1.0084 - val_dice_coef: 0.7977 - val_accuracy: 0.6452\n",
      "\n",
      "Epoch 00012: val_loss improved from 1.02250 to 1.00837, saving model to Dec19_noise10.h5\n",
      "Epoch 13/30\n",
      "9/9 [==============================] - 1529s 184s/step - loss: 0.9605 - dice_coef: 0.8078 - accuracy: 0.6750 - val_loss: 1.3475 - val_dice_coef: 0.7875 - val_accuracy: 0.6326\n",
      "\n",
      "Epoch 00013: val_loss did not improve from 1.00837\n",
      "Epoch 14/30\n",
      "9/9 [==============================] - 1525s 183s/step - loss: 1.0065 - dice_coef: 0.7973 - accuracy: 0.6454 - val_loss: 0.9527 - val_dice_coef: 0.8147 - val_accuracy: 0.6861\n",
      "\n",
      "Epoch 00014: val_loss improved from 1.00837 to 0.95272, saving model to Dec19_noise10.h5\n",
      "Epoch 15/30\n",
      "9/9 [==============================] - 1528s 184s/step - loss: 1.0052 - dice_coef: 0.7978 - accuracy: 0.6459 - val_loss: 0.9861 - val_dice_coef: 0.8061 - val_accuracy: 0.6675\n",
      "\n",
      "Epoch 00015: val_loss did not improve from 0.95272\n",
      "Epoch 16/30\n",
      "9/9 [==============================] - 1528s 184s/step - loss: 0.9698 - dice_coef: 0.8029 - accuracy: 0.6533 - val_loss: 0.9472 - val_dice_coef: 0.8127 - val_accuracy: 0.6804\n",
      "\n",
      "Epoch 00016: val_loss improved from 0.95272 to 0.94718, saving model to Dec19_noise10.h5\n",
      "Epoch 17/30\n",
      "9/9 [==============================] - 1511s 182s/step - loss: 0.8833 - dice_coef: 0.8258 - accuracy: 0.7223 - val_loss: 0.9361 - val_dice_coef: 0.8168 - val_accuracy: 0.6920\n",
      "\n",
      "Epoch 00017: val_loss improved from 0.94718 to 0.93614, saving model to Dec19_noise10.h5\n",
      "Epoch 18/30\n",
      "9/9 [==============================] - 1517s 183s/step - loss: 1.0211 - dice_coef: 0.7992 - accuracy: 0.6479 - val_loss: 0.9869 - val_dice_coef: 0.8074 - val_accuracy: 0.6712\n",
      "\n",
      "Epoch 00018: val_loss did not improve from 0.93614\n",
      "Epoch 19/30\n",
      "9/9 [==============================] - 1521s 183s/step - loss: 0.9491 - dice_coef: 0.8102 - accuracy: 0.6828 - val_loss: 0.9580 - val_dice_coef: 0.8172 - val_accuracy: 0.6923\n",
      "\n",
      "Epoch 00019: val_loss did not improve from 0.93614\n",
      "Epoch 20/30\n",
      "9/9 [==============================] - 1517s 183s/step - loss: 0.9885 - dice_coef: 0.8006 - accuracy: 0.6464 - val_loss: 0.9029 - val_dice_coef: 0.8225 - val_accuracy: 0.7033\n",
      "\n",
      "Epoch 00020: val_loss improved from 0.93614 to 0.90289, saving model to Dec19_noise10.h5\n",
      "Epoch 21/30\n",
      "9/9 [==============================] - 1521s 183s/step - loss: 0.9942 - dice_coef: 0.7988 - accuracy: 0.6500 - val_loss: 0.8932 - val_dice_coef: 0.8244 - val_accuracy: 0.7066\n",
      "\n",
      "Epoch 00021: val_loss improved from 0.90289 to 0.89318, saving model to Dec19_noise10.h5\n",
      "Epoch 22/30\n",
      "9/9 [==============================] - 1515s 182s/step - loss: 0.9549 - dice_coef: 0.8107 - accuracy: 0.6875 - val_loss: 0.8878 - val_dice_coef: 0.8305 - val_accuracy: 0.7134\n",
      "\n",
      "Epoch 00022: val_loss improved from 0.89318 to 0.88779, saving model to Dec19_noise10.h5\n",
      "Epoch 23/30\n",
      "9/9 [==============================] - 1515s 182s/step - loss: 0.9049 - dice_coef: 0.8210 - accuracy: 0.6931 - val_loss: 0.9046 - val_dice_coef: 0.8183 - val_accuracy: 0.6673\n",
      "\n",
      "Epoch 00023: val_loss did not improve from 0.88779\n",
      "Epoch 24/30\n",
      "9/9 [==============================] - 1516s 182s/step - loss: 0.9683 - dice_coef: 0.8048 - accuracy: 0.6603 - val_loss: 0.8979 - val_dice_coef: 0.8295 - val_accuracy: 0.7117\n",
      "\n",
      "Epoch 00024: val_loss did not improve from 0.88779\n",
      "Epoch 25/30\n",
      "9/9 [==============================] - 1521s 183s/step - loss: 0.9175 - dice_coef: 0.8209 - accuracy: 0.6926 - val_loss: 0.9115 - val_dice_coef: 0.8223 - val_accuracy: 0.6992\n",
      "\n",
      "Epoch 00025: val_loss did not improve from 0.88779\n",
      "Epoch 26/30\n",
      "9/9 [==============================] - 1517s 182s/step - loss: 0.8764 - dice_coef: 0.8312 - accuracy: 0.7226 - val_loss: 0.8641 - val_dice_coef: 0.8342 - val_accuracy: 0.7264\n",
      "\n",
      "Epoch 00026: val_loss improved from 0.88779 to 0.86406, saving model to Dec19_noise10.h5\n",
      "Epoch 27/30\n",
      "9/9 [==============================] - 1511s 182s/step - loss: 0.8451 - dice_coef: 0.8337 - accuracy: 0.7174 - val_loss: 0.8430 - val_dice_coef: 0.8394 - val_accuracy: 0.7305\n",
      "\n",
      "Epoch 00027: val_loss improved from 0.86406 to 0.84301, saving model to Dec19_noise10.h5\n",
      "Epoch 28/30\n",
      "9/9 [==============================] - 1520s 183s/step - loss: 0.8959 - dice_coef: 0.8215 - accuracy: 0.6985 - val_loss: 0.8699 - val_dice_coef: 0.8316 - val_accuracy: 0.7200\n",
      "\n",
      "Epoch 00028: val_loss did not improve from 0.84301\n",
      "Epoch 29/30\n",
      "9/9 [==============================] - 1522s 183s/step - loss: 0.8867 - dice_coef: 0.8229 - accuracy: 0.6896 - val_loss: 0.8719 - val_dice_coef: 0.8303 - val_accuracy: 0.6899\n",
      "\n",
      "Epoch 00029: val_loss did not improve from 0.84301\n",
      "Epoch 30/30\n",
      "9/9 [==============================] - 1518s 183s/step - loss: 0.8970 - dice_coef: 0.8221 - accuracy: 0.6900 - val_loss: 0.8523 - val_dice_coef: 0.8380 - val_accuracy: 0.7335\n",
      "\n",
      "Epoch 00030: val_loss did not improve from 0.84301\n"
     ]
    }
   ],
   "source": [
    "# Configure the model for training.\n",
    "# We use the \"sparse\" version of categorical_crossentropy\n",
    "# because our target data is integers.\n",
    "from keras.optimizers import Adam, SGD\n",
    "from keras_unet.metrics import iou, iou_thresholded\n",
    "#from keras_unet.losses import jaccard_distance\n",
    "\n",
    "from keras.callbacks import ModelCheckpoint\n",
    "\n",
    "\n",
    "val_len = val_gen.__len__()\n",
    "train_len = train_gen.__len__()\n",
    "\n",
    "steps_per_ep = train_len // batch_size\n",
    "\n",
    "model_filename = 'Dec19_noise10.h5'\n",
    "callback_checkpoint = ModelCheckpoint(\n",
    "    model_filename, \n",
    "    verbose=1, \n",
    "    monitor='val_loss', \n",
    "    save_best_only=True,\n",
    ")\n",
    "\n",
    "\n",
    "model.compile(optimizer=Adam( lr = 1e-4, decay=1e-5),\n",
    "              loss='categorical_crossentropy',\n",
    "              metrics= [dice_coef, 'accuracy']\n",
    "              )\n",
    "\n",
    "\n",
    "# Train the model, doing validation at the end of each epoch.\n",
    "epochs = 30\n",
    "history = model.fit(train_gen, \n",
    "                    epochs=epochs, \n",
    "                    steps_per_epoch=steps_per_ep, \n",
    "                    validation_data=val_gen, \n",
    "                    callbacks=[callback_checkpoint])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "worldwide-broadway",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best Loss :  0.8747829794883728\n",
      "Best Validation Loss :  0.8430138826370239\n",
      "Best Accuracy :  0.7148377299308777\n",
      "Best Validation Accuracy :  0.7335132360458374\n",
      "Best Dice Coef :  0.8285307884216309\n",
      "Best Validation Dice Coef :  0.8394444584846497\n"
     ]
    }
   ],
   "source": [
    "print(\"Best Loss : \", min(history.history['loss']))\n",
    "print(\"Best Validation Loss : \", min(history.history['val_loss']))\n",
    "\n",
    "print(\"Best Accuracy : \", max(history.history['accuracy']))\n",
    "print(\"Best Validation Accuracy : \", max(history.history['val_accuracy']))\n",
    "\n",
    "print(\"Best Dice Coef : \", max(history.history['dice_coef']))\n",
    "\n",
    "print(\"Best Validation Dice Coef : \", max(history.history['val_dice_coef']))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "photographic-gates",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "clinical-sperm",
   "metadata": {},
   "outputs": [],
   "source": [
    "val_preds = model.predict(val_gen)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "blind-bloom",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "id": "novel-physiology",
   "metadata": {},
   "outputs": [],
   "source": [
    "def read_predicted_mask(img):\n",
    "    img = np.array(img)#img = np.array(Image.fromarray(img.astype(np.uint8)).resize((256, 256)))\n",
    "    img = np.argmax(img, axis=-1)\n",
    "    img = np.expand_dims(img, axis=-1)\n",
    "    img = np.asarray(img, dtype=np.uint8)\n",
    "    print(img.shape)\n",
    "    return img\n",
    "\n",
    "# land - gree\n",
    "# high - red\n",
    "# moderate - yellow\n",
    "# low - black\n",
    "# water - blue\n",
    "\n",
    "def make_colour_mask_pred(segment_mask):\n",
    "    #print(segment_mask.shape) \n",
    "    print(np.unique(segment_mask))\n",
    "    colour_table = np.zeros((256, 1, 3), dtype=np.uint8)\n",
    "    colour_table[3] = [0,128,0] \n",
    "    colour_table[1] = [255,0,0]\n",
    "    colour_table[2] = [255,255,0]\n",
    "    colour_table[0] = [0,0,0]\n",
    "    colour_table[4] = [0,0,255]\n",
    "\n",
    "    colour_mask = cv2.applyColorMap(segment_mask, colour_table)\n",
    "    plt.imshow(colour_mask)\n",
    "    return colour_mask"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "nuclear-placement",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(256, 256, 1) read mask\n",
      "[3 4]\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAQYAAAD8CAYAAACVSwr3AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuNCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8QVMy6AAAACXBIWXMAAAsTAAALEwEAmpwYAAAXA0lEQVR4nO3de5RcVZXH8e9OSIKGRxICMQQwgGFGQAkxIigqCigwjoFREQYkKhpUVEDGWVHHGV+IusD3MwxIYBQGFRZxeChkXKKzBElYGBIQCRIkIQ8BefkgJNnzx7kNlVR3uqq7bu1z6/4+a9Xq6tvVVb+udO+ce+55mLsjItJoRHQAEcmPCoOINFFhEJEmKgwi0kSFQUSaqDCISJPSCoOZHWVmd5vZcjObW9briEjnWRnjGMxsJPA74EhgJXArcKK739nxFxORjiurxXAQsNzdf+/u64HLgVklvZaIdNg2JT3vFOCBhs9XAi8b6MFmpuGX/TJgcnGLchugf54e8ZC779zKA8sqDIMysznAnKjXz9+2wJnAuYEZvMixPjCDdND9rT6wrMKwCti94fPdimPPcPd5wDxQi6F/k4ktCgAXAxuDM0iEsvoYbgWmmdmeZjYaOAFYUNJr9aCRwOejQwDnoMJQT6W0GNx9g5m9H/gJ6bf8IndfVsZr9abrgCOiQ0iNldbH4O7XAteW9fy9axQwndTxGOkE4L7gDBJFIx+zsgtwMzAxOgjwJLApOoQECbsqIf35NDAjOgRwF7A2OoQEUoshGy8BDowOUbgUWBQdQgKpxZCFF5Ku3ObQWhBRiyETE8mnKFwHXBAdQoKpMITbm3RVNwcbgTXAQ9FBJJhOJUKNIHX0jYoOUrgTeGd0CMmAWgyh3oj+CSRH+q0McyownzQwNAdPAV+KDiGZUGEI81Zgh+gQDdYD340OIZlQYQjxMeDQ6BBb0ARXeZY6H7vuZOCT5HMK0WdqdADJiFoMXTUW2If8igLA09EBJCNqMXTNaOAjpNOI3PwQFQZpVMoq0W2HqMUKTuOAP0WHGMCLgKXRIaR8i919ZisP1KlE11wZHUCkZTqV6IpfAK+IDjGA9wC/jQ4hmVGLoXQTSQu7Rq/INJBHgQ3RISQzKgyl2gu4ijRRKkf3kSZNiWxOhaFUp5DfQKZGlwE/jw4hGVJhKM1LgGOjQ4gMiTofSzEV+G/yPYWA1CH6jegQkim1GEoxjryLAsBjwIPRISRTKgwdtz9pI9icrWeLHQNFNqPC0HG/JN9Lk33+QBq/INI/FYaOmgOMiQ4xiA3A16NDSOZUGDrmg6SNaLeNDjKIjcBXokNI5lQYOualpE7H3B0WHUAqQJcrh20b4GzSJrC5exL4XXQIqQAVhmF7DfC56BAtOgJ4JDqEVIBOJYZlLHB4dIgW/QptJCOtUothyEYA51Gdy34XAvdGh5CKUIthyEaQLk9WwdXA9dEhpEJUGIbEgJvJfyATpGXh70cjHaUdwzqVMLMVwBOki+Mb3H2mmU0gzSCaCqwAjnf3XBc7HKL/Je1OXYXCcBNwVnQIqZhOtBhe4+7TGxaZnAssdPdpwMLi8x6yNzCBahSFp4AlwKboIFIxZZxKzCJtykjx8dgSXiPQ2cCLo0O06GHSiEyR9gy3MDjwUzNbbGZ9PXGT3H11cX8NMKm/bzSzOWa2yMwWDTNDF70SeFV0iBZtAj4aHUIqalj7SpjZFHdfZWa7ADcAHwAWuPu4hsf8yd3HD/I8FdhX4kWkJeBfEB2kRRtJE7o2RgeRfHRnXwl3X1V8XEda9fQgYK2ZTQYoPq4bzmvkYwLVKQobSJ2jKgoyNEMuDGY21sy277sPvI60ndECYHbxsNmki+gVN4N0JaIqjiJ1OooMzXAuV04CrjKzvuf5vrtfb2a3AleY2amkC+jHDz9mpL4xC1UZ8vFbNPRZhmvIhcHdfw8c0M/xh6nOBIIWvJvqFAWAbwO/iQ4hFVel3/gAHwbOJ89t6/tzE/CT6BDSA1QYtupwYLvoEG24F+1DKZ2gwjCgTwGvjQ7RhsXA6dEhpEeoMPRrB2AKMCo6SIs2kloLf40OIj1ChaHJjqTWwjujg7ThMeCt0SGkh6gwNNkTOCM6RBuctGCMSOeoMGzmucAXo0MMQRUzS860tNtmbibNiagKJ41yXB8dRHqMWgzP2J78N6Ld0ltIc9cqMAdNKkWF4Rm3kE4lqmIV8EdUFKQMKgwAvJp0NaIqVgDvJY10FOk8FQaOBC4Ado0O0qI1pGUvfhwdRHqYCgMHA9OiQ7ThYeB/okNIj6t5YTgK+FB0iDY8SdpmTqRcNb5cuQ/pf96qzJx8iNSyeTQ4h9RBTVsMI0jb1lelKECa6flodAipiZoWhtHApdEh2vBTUt+CSHfUtDB8ITpAG64H3oe2mJNuqmkfw8lUYyepXwGnAX+IDiI1U8MWww1UYzCTk0Y2qihI99WsMEwkLW6d+4/twC9Iu/2JdF/ufyEddh7VmD35NHBYdAipsRoVhpcC+0eHaNE30OQoiVSTzsf9SfstzIgO0oL/AD4fHUJqriYthudRjaLwKdLpzlPRQaTmalAYplCN7TOfJk2n/ktwDpFanEr8HfkvwPJn4Fzgu9FBRICebzHMIo1byNlfSacP50QHEXmGucf3fptZSSEeBCaX89Qds5rqLBIjFbfY3We28sAebjF8FhgfHWIQm4BTo0OINOnRPoZ/B84Etg3OsTVOGsT0i+AcIs16sMUwBtgdeE50kEG8ARUFyVUPFoZTgHdFhxjEH0grMonkadDCYGYXmdk6M1vacGyCmd1gZvcUH8cXx83Mvmpmy81siZl1eVTRzsDLu/uSbVsOvBv4dXQQkQG10mK4mLRqaqO5wEJ3nwYsLD4HOJq0MOE0YA7wrc7EbMVY4GvA27v3km17ADidtCKTSMbcfdAbMBVY2vD53cDk4v5k4O7i/neAE/t73CDP78O/TXTwjG+PO7y0Az+nbroN+baolb93dx9yH8Mkd19d3F9DWuQA0vjjBxoet7I4VrKRwKLyX2ZY1gO3RocQacmwL1e6uw9lgJKZzSGdbnTAXcAenXmqUqwF9ooOIdKyobYY1prZZIDi47ri+CrStcI+uzHAKqbuPs/dZ7Y6EmtgLwK2I981HJeS1oLQ5CipjqEWhgXA7OL+bJ6dvrgAOKW4OnEw8FjDKUdJziXfYc83Ayex+dmVSP4GPZUws8tIQ/QmmtlK0koinwOuMLNTgfuB44uHXwscQ7om9xfgHSVkbvAm4MXlvsSwXAcsiQ4h0rYKT6J6HWmacq4TkH4GvA3tByEZaXkSVUXnShjpQkiORcGBZaQp308EZxHpM5p0Zaw1FRwSbaSzlfnRQQZwF6lDVEVBcrEr6Yy/dRUsDNuRdqnO9SrEIdEBRBrsR+rrel5b31XBU4kPRgfYigvRQq6Sh11JFwzfwFA66CtYGOYO/pAQXwc+jgqDxNuONMXpyCE/Q8UKw5XkubDrxaSi8GhsDBEAfgkcMKxnqFAfw3zgjeQXeQNpSsijwTlEtgXuoBNje3L7KxvAzqTOk5HRQbawntSv8PHoIFJ7k4DrSbuuDb9jviKnEnNIA5pysgmYB3wgOojU2o6k5VJOAF7dsWetQGH4e4bTiVKec0iLzopEGU3ak6TzSxlWoDC8gE5Wws75enQAqb3LgeNKeeYKFIbcBjI5qdn2cHQQqbVraF5xsXMy73zcG/hRdIgGfyP1d/wA2BicRerpOcAVpKJQ3p9vxoXBgAOBUdFBCk+QZpz/J6nVINJt44DzgbdQ9p9uxtOux5A2fM3lVOKXwCujQ0htbQ98huFNCbBemHb9megADZ4AvhwdQmprJGlrhNmDPbBjMi4M7ySP1sIm0kSUm6KDSG1dRfod7J5MC8M1pPOpHByCdo2SGCNIy6geTbf/k8yw83E8acpoDtEeAh6MDiG1NBa4hLQoUff/FnL469vC+cD06BDAPaRKvTI6iNTOONLq5ycRdTqdWWGYDuwbHYK0PNtp5L+7lfSescCniZ6Dk1kfw6uBl0WHABaTVnkW6SYjTcz75+ggORWGmcCHokOQ9uHV5CiJ8GNSn0K8jArDTsTvP/lH4CDg8eAcUi/bkIb+d//qw0Ay6WMYRVrJNtIKUmFSUZBu2gG4APhHsvlzJJskOxBfKY8mTZIS6ZZxpI7GtxP/+7+5TE4lpga//g/RNGrprm1JlyTfEx2kX5kUhkg/BM4k9S+IdMslpFmSeap5YVgIvA8VBemuchdZ6YRM+hgibATWoqIg3TOadPWh3EVWOiHvdKXZRNr/8qToIFIbO5KmTv8TVfizyz9hKS4Hjo0OIbWxI/BJ0rKA1VDDwvA10loPIt0wijQx8IzoIG2pYefjfLTxrHTPFcCs6BBtG7TFYGYXmdk6M1vacOwTZrbKzG4vbsc0fO0jZrbczO42s9eXFbx9DpxN2ttPpBuuIe23mtfgpVa0cipxMf1fW/mSu08vbtcCmNm+pE0X9iu+55tmlsGGk+uBz5LWbVwfG0VqYj5VuPowkEFTu/tNwCMtPt8s4HJ3f8rd7wOWk2YlBfs58G+kqxEiZduVtAlzNYsCDC/5+81sSXGqMb44NgV4oOExK4tjTcxsjpktMrNF5Y4leBK4tsTnF2m0O6mDO7dNmNsz1MLwLdI2UdOB1aRu17a4+zx3n5nWud95iDEGs5G0Es6XS3p+kUY7Ad8kjVWotiFdlXD3tX33zewC0mghgFWkktlnt+JYkDcBV8e9vNTINqTOxhxWIBu+IbUYzGxyw6fHAX1XLBYAJ5jZGDPbE5hG2Nrrx5FWxBHphtvJojutQwZtMZjZZcBhwEQzW0nawPEwM5tOuga4grRyKu6+zMyuAO4ENgCnu3vA7q+PkJZ9V2ejdMMupM7G6l2WHEgme1fO9M6tyLySNGNSrQXphheSdj/fLzpIC1rfu7K611P69SBwFioK0h0HABdSjaLQnh4aEv048C7i146UetgH+A690tm4pR4pDA68lrQfhEg3TKJXiwL0zKnE34Al0SGkNvYgXZrsXT1QGFaSmnVPRweRWngBcC+wfXSQUlW8MCwjTc/QxrPSLQvpmTPwrahwYbiDNHzituggUhsnk1Zj6n0VLwz/Fx1CamM2cB51KQwVbBM5aePZD0cHkdqYRZonuFN0kK6pYGF4EDgQbScn3TGCNNy5PkUBKncqcRepV1hFQbphJGlBsm9HB+m6ihWGN6OiIN3zduB70SFCVKgwXEraOUqkGz5MHVsKfSpSGL5P+ofSjtTSLadSyS64Dsn8J3fgRuB04NHYKFIj3wb2ig4RKuPC4MAtwNGktRtFyjYG+BTwbirTmC5JxoXhKeCQ6BBSK/8A/Gt0iCxkXBYviQ4gtTKOqi/53kmZFobPAR+MDiG1MYq0I8Jp0UGykeGpxGeBc9HGs9I91wBHRIfISkaFYRNpqaxPo0FM0j03kFb/6p0Vnjshk8LgwOWk1Z1FumUiMJlsz6gDZfKOPA2cFB1CamUq6T+j3lvhuRMyKQxlbmor0p83A4dHh8hWJoVhTXQAqZX9gROjQ2Qtk8Ig0k2TgRnRIbKmwiA1sztwVXSI7KkwSM1MA8ZGh8ieCoPUyDakcQsyGBUGqZGz0UCm1qgwSI2chQpDa1QYRKSJCoPUxI3ALtEhKmPQwmBmu5vZz8zsTjNbZmZnFMcnmNkNZnZP8XF8cdzM7KtmttzMlpiZLhhLsD1I+0LoNKJVrbQYNgBnu/u+wMHA6Wa2LzAXWOju00g7fc4tHn806ZrQNGAOaaK7SKC5wPToEJUyaGFw99Xufltx/wnSri9TSPt2zS8eNh84trg/C7jEk5uBcWY2udPBRVpzCPDy6BCV01Yfg5lNJe0Pdwswyd1XF19aA0wq7k8BHmj4tpXFMZEABxQ3aUfL6zGY2XbAj4Az3f1xs2fP19zdzczbeWEzm0M61RApyUzgnOgQldRSi8HMRpGKwvfc/cri8Nq+U4Ti47ri+CrSgPQ+uxXHNuPu89x9prvPHGp4kYGNAHYFJkQHqaRWrkoYcCFwl7t/seFLC4DZxf3ZwNUNx08prk4cDDzWcMoh0iUTefZXUtrVyqnEK4C3AXeY2e3FsY+SlnK+wsxOBe4Hji++di1wDLAc+Avwjk4GFmnN7MEfIgMy97a6BsoJ0Wb/hMjWGWmV8VHRQTJji1s9ddfIRxFposIgPehWslkAvaJUGKQHTUDDn4dHhUF6zEzgudEhKk+FQXrMZ3h2EK4MlQqDiDRRYZAechZp0pQMlwqD9JDJwA7RIXqCCoP0iB1Jw6ClE1QYpEe8AY2+7xwVBhFposIgPWAv4EPRIXqKCoP0gB3QJrWdpcIgIk1UGKTiRgL7RYfoOSoMUnFjgf+KDtFzVBhEpIkKg1TcD6ID9CQVBqm4Q6MD9CQVBqmwndGCLOVQYZAK+znwnOgQPUmFQUSaqDBIRb2XNM1ayqDCIBV1ODAuOkTPUmGQCjqbNM1ayqLCIBX0XGBMdIiepsIgFTMJ2Cc6RM9TYZCKOQQ4OTpEz1NhEJEmKgxSIVOBLwdnqAcVBqmQ0cDzo0PUggqDVIShTsfuUWGQihgDLIgOURsqDCLSZNDCYGa7m9nPzOxOM1tmZmcUxz9hZqvM7PbidkzD93zEzJab2d1m9voyfwCpi0uiA9TKNi08ZgNwtrvfZmbbA4vN7Ibia19y9/MaH2xm+wInkFbo3BW40cz2cfeNnQwudXMEWnuhewZtMbj7ane/rbj/BHAXMGUr3zILuNzdn3L3+4DlwEGdCCt1tSMqCt3VVh+DmU0FDgRuKQ6938yWmNlFZja+ODYFeKDh21bSTyExszlmtsjMFrUfW+rlRjSTsrtaLgxmth3wI+BMd38c+BawNzAdWA2c384Lu/s8d5/p7jPb+T4RKV9LhcHMRpGKwvfc/UoAd1/r7hvdfRNwAc+eLqwCdm/49t2KYyJDcBKwR3SI2mnlqoQBFwJ3ufsXG443Lp9zHLC0uL8AOMHMxpjZnsA04Nediyz1ciSwS3SI2mnlqsQrgLcBd5jZ7cWxjwInmtl0wIEVwGkA7r7MzK4A7iRd0ThdVyREqsXcPToDZvZH4M/AQ9FZWjCRauSE6mRVzs7rL+vz3X3nVr45i8IAYGaLqtARWZWcUJ2sytl5w82qIdEi0kSFQUSa5FQY5kUHaFFVckJ1sipn5w0razZ9DCKSj5xaDCKSifDCYGZHFdOzl5vZ3Og8WzKzFWZ2RzG1fFFxbIKZ3WBm9xQfxw/2PCXkusjM1pnZ0oZj/eay5KvFe7zEzGZkkDW7aftbWWIgq/e1K0shuHvYDRgJ3AvsRVrQ7zfAvpGZ+sm4Api4xbEvAHOL+3OBzwfkehUwA1g6WC7gGOA60hTFg4FbMsj6CeBf+nnsvsXvwRhgz+L3Y2SXck4GZhT3twd+V+TJ6n3dSs6OvafRLYaDgOXu/nt3Xw9cTpq2nbtZwPzi/nzg2G4HcPebgEe2ODxQrlnAJZ7cDIzbYkh7qQbIOpCwafs+8BIDWb2vW8k5kLbf0+jC0NIU7WAO/NTMFpvZnOLYJHdfXdxfQ9oeKQcD5cr1fR7ytP2ybbHEQLbvayeXQmgUXRiq4FB3nwEcDZxuZq9q/KKntlp2l3ZyzdVgWNP2y9TPEgPPyOl97fRSCI2iC0P2U7TdfVXxcR1wFakJtravyVh8XBeXcDMD5cruffZMp+33t8QAGb6vZS+FEF0YbgWmmdmeZjaatFZkNmuEm9nYYp1LzGws8DrS9PIFwOziYbOBq2MSNhko1wLglKIX/WDgsYamcYgcp+0PtMQAmb2vA+Xs6HvajV7UQXpYjyH1qt4LfCw6zxbZ9iL15v4GWNaXD9gJWAjcQ1p3bEJAtstIzcWnSeeMpw6Ui9Rr/o3iPb4DmJlB1kuLLEuKX9zJDY//WJH1buDoLuY8lHSasAS4vbgdk9v7upWcHXtPNfJRRJpEn0qISIZUGESkiQqDiDRRYRCRJioMItJEhUFEmqgwiEgTFQYRafL/rh9iLy7JlDUAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import PIL as pl\n",
    "from PIL import ImageOps\n",
    "def display_mask(i):\n",
    "    \"\"\"Quick utility to display a model's prediction.\"\"\"\n",
    "    print(i)\n",
    "    mask = np.argmax(val_preds[i], axis=-1)\n",
    "    mask = np.array(mask)\n",
    "    mask = np.array(Image.fromarray(mask.astype(np.uint8)).resize((256, 256)))\n",
    "    mask = np.expand_dims(mask, axis=-1)\n",
    "    return mask\n",
    "\n",
    "# Display results for validation image #10\n",
    "index = 120\n",
    "\n",
    "# Display input image\n",
    "fp = val_input_img_paths[index]\n",
    "img = rasterio.open(fp)\n",
    "show(img)\n",
    "\n",
    "# Display ground-truth target mask\n",
    "file = val_target_img_paths[index]\n",
    "im = read_original_mask(file)\n",
    "im = make_colour_mask_pred(im)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "conscious-reggae",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(256, 256, 1)\n",
      "[0 2 3 4]\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"
    }
   ],
   "source": [
    "# Display mask predicted by our model\n",
    "im = read_predicted_mask(val_preds[120])\n",
    "#print(im)\n",
    "prediction = make_colour_mask_pred(im)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "abstract-legend",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(256, 256, 1)\n",
      "[0 1 2 3 4]\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"
    }
   ],
   "source": [
    "# Display mask predicted by our model\n",
    "im = read_predicted_mask(val_preds[550])\n",
    "#print(im)\n",
    "prediction = make_colour_mask_pred(im)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "suspected-sperm",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(256, 256, 1)\n",
      "[0 1 2 3 4]\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"
    }
   ],
   "source": [
    "# Display mask predicted by our model\n",
    "im = read_predicted_mask(val_preds[150])\n",
    "#print(im)\n",
    "prediction = make_colour_mask_pred(im)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "narrative-plaintiff",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(256, 256, 1)\n",
      "[0 2 3 4]\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"
    }
   ],
   "source": [
    "# Display mask predicted by our model\n",
    "im = read_predicted_mask(val_preds[525])\n",
    "#print(im)\n",
    "prediction = make_colour_mask_pred(im)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "popular-chemistry",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(256, 256, 1)\n",
      "[0 1 2 3 4]\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"
    }
   ],
   "source": [
    "# Display mask predicted by our model\n",
    "im = read_predicted_mask(val_preds[122])\n",
    "#print(im)\n",
    "prediction = make_colour_mask_pred(im)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "norwegian-friendship",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(256, 256, 1)\n",
      "[0 1 2 3 4]\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"
    }
   ],
   "source": [
    "# Display mask predicted by our model\n",
    "im = read_predicted_mask(val_preds[142])\n",
    "#print(im)\n",
    "prediction = make_colour_mask_pred(im)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "id": "individual-mobility",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(256, 256, 1)\n",
      "[0 3 4]\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"
    }
   ],
   "source": [
    "# Display mask predicted by our model\n",
    "im = read_predicted_mask(val_preds[1172])\n",
    "#print(im)\n",
    "prediction = make_colour_mask_pred(im)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "id": "southeast-mount",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(256, 256, 1)\n",
      "[0 1 2 3 4]\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"
    }
   ],
   "source": [
    "# Display mask predicted by our model\n",
    "im = read_predicted_mask(val_preds[512])\n",
    "#print(im)\n",
    "prediction = make_colour_mask_pred(im)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "id": "ancient-hours",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(256, 256, 1)\n",
      "[0 1 2 3 4]\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"
    }
   ],
   "source": [
    "# Display mask predicted by our model\n",
    "im = read_predicted_mask(val_preds[122])\n",
    "#print(im)\n",
    "prediction = make_colour_mask_pred(im)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "southwest-blair",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}