tutorial119_multiclass_semantic_segmentation.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "tutorial119_multiclass_semantic_segmentation.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "mount_file_id": "1RvuTk-Scq7viDQTc1Y2kDMNsD1xHXFyK",
      "authorship_tag": "ABX9TyNh0U+v2EtxiJ6GesFrR7ku",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/bnsreenu/python_for_image_processing_APEER/blob/master/tutorial119_multiclass_semantic_segmentation.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "5jWYe6YIa1Ie"
      },
      "source": [
        "https://youtu.be/zpYq1gr1910"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "SscH4lze36iJ"
      },
      "source": [
        "**Multiclass semantic segmentation**\n",
        "\n",
        "\n",
        "The dataset used in this video can be downloaded from the link below. This dataset can be used to train and test machine learning algorithms designed for multiclass semantic segmentation. Please read the Readme document for more information. \n",
        "https://drive.google.com/file/d/1HWtBaSa-LTyAMgf2uaz1T9o1sTWDBajU/view?usp=sharing\n",
        "\n",
        "To annotate your own images, please visit www.apeer.com"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "I0NRq8jOY5bX"
      },
      "source": [
        "#from keras.utils import normalize\n",
        "import os\n",
        "import glob\n",
        "import cv2\n",
        "import numpy as np\n",
        "from matplotlib import pyplot as plt"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3MJ7xjSDZDJZ"
      },
      "source": [
        "#Resizing images, if needed\n",
        "SIZE_X = 128 \n",
        "SIZE_Y = 128\n",
        "n_classes=4 #Number of classes for segmentation\n",
        "num_images = 200  #Total 1600 available"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "e-qxgNa8ZQ_H"
      },
      "source": [
        "image_names = glob.glob(\"/content/drive/MyDrive/Colab Notebooks/data/sandstone/128_patches/images/*.tif\")\n",
        "image_names.sort()\n",
        "image_names_subset = image_names[0:num_images]\n",
        "images = [cv2.imread(image, 0) for image in image_names_subset]\n",
        "image_dataset = np.array(images)\n",
        "image_dataset = np.expand_dims(image_dataset, axis = 3)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JMVZCMFPZGPX"
      },
      "source": [
        "mask_names = glob.glob(\"/content/drive/MyDrive/Colab Notebooks/data/sandstone/128_patches/masks/*.tif\")\n",
        "mask_names.sort()\n",
        "mask_names_subset = mask_names[0:num_images]\n",
        "masks = [cv2.imread(mask, 0) for mask in mask_names_subset]\n",
        "mask_dataset = np.array(masks)\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "QzAouaVaZoVg",
        "outputId": "bf942f6f-1ff5-448b-beaa-44f01c523110"
      },
      "source": [
        "print(\"Image data shape is: \", image_dataset.shape)\n",
        "print(\"Mask data shape is: \", mask_dataset.shape)\n",
        "print(\"Max pixel value in image is: \", image_dataset.max())\n",
        "print(\"Labels in the mask are : \", np.unique(mask_dataset))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Image data shape is:  (200, 128, 128, 1)\n",
            "Mask data shape is:  (200, 128, 128)\n",
            "Max pixel value in image is:  255\n",
            "Labels in the mask are :  [1 2 3 4]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ibHIDZGfZkTh",
        "outputId": "7a4e6ce5-7df7-49e0-e519-ab64f0ba1ec2"
      },
      "source": [
        "#Encode labels to 0, 1, 2, 3, ... but multi dim array so need to flatten, encode and reshape\n",
        "from sklearn.preprocessing import LabelEncoder\n",
        "labelencoder = LabelEncoder()\n",
        "n, h, w = mask_dataset.shape  \n",
        "mask_dataset_reshaped = mask_dataset.reshape(-1,1)\n",
        "mask_dataset_reshaped_encoded = labelencoder.fit_transform(mask_dataset_reshaped)\n",
        "mask_dataset_encoded = mask_dataset_reshaped_encoded.reshape(n, h, w)\n",
        "\n",
        "np.unique(mask_dataset_encoded)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/sklearn/preprocessing/_label.py:251: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
            "  y = column_or_1d(y, warn=True)\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([0, 1, 2, 3])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 6
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "QDQ_8v3FhENA",
        "outputId": "af8a0fa2-0fac-43b5-ca97-5098f44d5e95"
      },
      "source": [
        "mask_dataset_encoded = np.expand_dims(mask_dataset_encoded, axis = 3)\n",
        "print(mask_dataset_encoded.shape)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(200, 128, 128, 1)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9SgFwOThZs_A"
      },
      "source": [
        "#Normalize images\n",
        "image_dataset = image_dataset /255.  #Can also normalize or scale using MinMax scaler"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IMZPM7edZ4nN"
      },
      "source": [
        "#Split training data\n",
        "from sklearn.model_selection import train_test_split\n",
        "X_train, X_test, y_train, y_test = train_test_split(image_dataset, mask_dataset_encoded, test_size = 0.2, random_state = 42)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Yh2fI72laFn1"
      },
      "source": [
        "from tensorflow.keras.utils import to_categorical\n",
        "train_masks_cat = to_categorical(y_train, num_classes=n_classes)\n",
        "y_train_cat = train_masks_cat.reshape((y_train.shape[0], y_train.shape[1], y_train.shape[2], n_classes))\n",
        "\n",
        "test_masks_cat = to_categorical(y_test, num_classes=n_classes)\n",
        "y_test_cat = test_masks_cat.reshape((y_test.shape[0], y_test.shape[1], y_test.shape[2], n_classes))\n",
        "\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "CDMoprb5hrvY",
        "outputId": "07ba7bf8-bdad-4648-9b3e-d337b3c6f95c"
      },
      "source": [
        "print(y_train_cat.shape)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(160, 128, 128, 4)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CNEIG7sKaR0D"
      },
      "source": [
        "# Building Unet by dividing encoder and decoder into blocks\n",
        "\n",
        "from keras.models import Model\n",
        "from keras.layers import Input, Conv2D, MaxPooling2D, UpSampling2D, concatenate, Conv2DTranspose, BatchNormalization, Dropout, Lambda\n",
        "from keras.optimizers import Adam\n",
        "from keras.layers import Activation, MaxPool2D, Concatenate\n",
        "\n",
        "\n",
        "def conv_block(input, num_filters):\n",
        "    x = Conv2D(num_filters, 3, padding=\"same\")(input)\n",
        "    x = BatchNormalization()(x)   #Not in the original network. \n",
        "    x = Activation(\"relu\")(x)\n",
        "\n",
        "    x = Conv2D(num_filters, 3, padding=\"same\")(x)\n",
        "    x = BatchNormalization()(x)  #Not in the original network\n",
        "    x = Activation(\"relu\")(x)\n",
        "\n",
        "    return x\n",
        "\n",
        "#Encoder block: Conv block followed by maxpooling\n",
        "\n",
        "\n",
        "def encoder_block(input, num_filters):\n",
        "    x = conv_block(input, num_filters)\n",
        "    p = MaxPool2D((2, 2))(x)\n",
        "    return x, p   \n",
        "\n",
        "#Decoder block\n",
        "#skip features gets input from encoder for concatenation\n",
        "\n",
        "def decoder_block(input, skip_features, num_filters):\n",
        "    x = Conv2DTranspose(num_filters, (2, 2), strides=2, padding=\"same\")(input)\n",
        "    x = Concatenate()([x, skip_features])\n",
        "    x = conv_block(x, num_filters)\n",
        "    return x\n",
        "\n",
        "#Build Unet using the blocks\n",
        "def build_unet(input_shape, n_classes):\n",
        "    inputs = Input(input_shape)\n",
        "\n",
        "    s1, p1 = encoder_block(inputs, 64)\n",
        "    s2, p2 = encoder_block(p1, 128)\n",
        "    s3, p3 = encoder_block(p2, 256)\n",
        "    s4, p4 = encoder_block(p3, 512)\n",
        "\n",
        "    b1 = conv_block(p4, 1024) #Bridge\n",
        "\n",
        "    d1 = decoder_block(b1, s4, 512)\n",
        "    d2 = decoder_block(d1, s3, 256)\n",
        "    d3 = decoder_block(d2, s2, 128)\n",
        "    d4 = decoder_block(d3, s1, 64)\n",
        "\n",
        "    if n_classes == 1:  #Binary\n",
        "      activation = 'sigmoid'\n",
        "    else:\n",
        "      activation = 'softmax'\n",
        "\n",
        "    outputs = Conv2D(n_classes, 1, padding=\"same\", activation=activation)(d4)  #Change the activation based on n_classes\n",
        "    print(activation)\n",
        "\n",
        "    model = Model(inputs, outputs, name=\"U-Net\")\n",
        "    return model"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "o3IGtgOAaJcC"
      },
      "source": [
        "IMG_HEIGHT = X_train.shape[1]\n",
        "IMG_WIDTH  = X_train.shape[2]\n",
        "IMG_CHANNELS = X_train.shape[3]\n",
        "input_shape = (IMG_HEIGHT, IMG_WIDTH, IMG_CHANNELS)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "HcQp3wxCaTZw",
        "outputId": "eaa198f6-cc50-4aa6-a635-bedff8c0e71e"
      },
      "source": [
        "model = build_unet(input_shape, n_classes=4)\n",
        "model.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy'])\n",
        "model.summary()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "softmax\n",
            "Model: \"U-Net\"\n",
            "__________________________________________________________________________________________________\n",
            "Layer (type)                    Output Shape         Param #     Connected to                     \n",
            "==================================================================================================\n",
            "input_1 (InputLayer)            [(None, 128, 128, 1) 0                                            \n",
            "__________________________________________________________________________________________________\n",
            "conv2d (Conv2D)                 (None, 128, 128, 64) 640         input_1[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization (BatchNorma (None, 128, 128, 64) 256         conv2d[0][0]                     \n",
            "__________________________________________________________________________________________________\n",
            "activation (Activation)         (None, 128, 128, 64) 0           batch_normalization[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_1 (Conv2D)               (None, 128, 128, 64) 36928       activation[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_1 (BatchNor (None, 128, 128, 64) 256         conv2d_1[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "activation_1 (Activation)       (None, 128, 128, 64) 0           batch_normalization_1[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "max_pooling2d (MaxPooling2D)    (None, 64, 64, 64)   0           activation_1[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_2 (Conv2D)               (None, 64, 64, 128)  73856       max_pooling2d[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_2 (BatchNor (None, 64, 64, 128)  512         conv2d_2[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "activation_2 (Activation)       (None, 64, 64, 128)  0           batch_normalization_2[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_3 (Conv2D)               (None, 64, 64, 128)  147584      activation_2[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_3 (BatchNor (None, 64, 64, 128)  512         conv2d_3[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "activation_3 (Activation)       (None, 64, 64, 128)  0           batch_normalization_3[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "max_pooling2d_1 (MaxPooling2D)  (None, 32, 32, 128)  0           activation_3[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_4 (Conv2D)               (None, 32, 32, 256)  295168      max_pooling2d_1[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_4 (BatchNor (None, 32, 32, 256)  1024        conv2d_4[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "activation_4 (Activation)       (None, 32, 32, 256)  0           batch_normalization_4[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_5 (Conv2D)               (None, 32, 32, 256)  590080      activation_4[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_5 (BatchNor (None, 32, 32, 256)  1024        conv2d_5[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "activation_5 (Activation)       (None, 32, 32, 256)  0           batch_normalization_5[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "max_pooling2d_2 (MaxPooling2D)  (None, 16, 16, 256)  0           activation_5[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_6 (Conv2D)               (None, 16, 16, 512)  1180160     max_pooling2d_2[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_6 (BatchNor (None, 16, 16, 512)  2048        conv2d_6[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "activation_6 (Activation)       (None, 16, 16, 512)  0           batch_normalization_6[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_7 (Conv2D)               (None, 16, 16, 512)  2359808     activation_6[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_7 (BatchNor (None, 16, 16, 512)  2048        conv2d_7[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "activation_7 (Activation)       (None, 16, 16, 512)  0           batch_normalization_7[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "max_pooling2d_3 (MaxPooling2D)  (None, 8, 8, 512)    0           activation_7[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_8 (Conv2D)               (None, 8, 8, 1024)   4719616     max_pooling2d_3[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_8 (BatchNor (None, 8, 8, 1024)   4096        conv2d_8[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "activation_8 (Activation)       (None, 8, 8, 1024)   0           batch_normalization_8[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_9 (Conv2D)               (None, 8, 8, 1024)   9438208     activation_8[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_9 (BatchNor (None, 8, 8, 1024)   4096        conv2d_9[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "activation_9 (Activation)       (None, 8, 8, 1024)   0           batch_normalization_9[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_transpose (Conv2DTranspo (None, 16, 16, 512)  2097664     activation_9[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "concatenate (Concatenate)       (None, 16, 16, 1024) 0           conv2d_transpose[0][0]           \n",
            "                                                                 activation_7[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_10 (Conv2D)              (None, 16, 16, 512)  4719104     concatenate[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_10 (BatchNo (None, 16, 16, 512)  2048        conv2d_10[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "activation_10 (Activation)      (None, 16, 16, 512)  0           batch_normalization_10[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_11 (Conv2D)              (None, 16, 16, 512)  2359808     activation_10[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_11 (BatchNo (None, 16, 16, 512)  2048        conv2d_11[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "activation_11 (Activation)      (None, 16, 16, 512)  0           batch_normalization_11[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_transpose_1 (Conv2DTrans (None, 32, 32, 256)  524544      activation_11[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "concatenate_1 (Concatenate)     (None, 32, 32, 512)  0           conv2d_transpose_1[0][0]         \n",
            "                                                                 activation_5[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_12 (Conv2D)              (None, 32, 32, 256)  1179904     concatenate_1[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_12 (BatchNo (None, 32, 32, 256)  1024        conv2d_12[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "activation_12 (Activation)      (None, 32, 32, 256)  0           batch_normalization_12[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_13 (Conv2D)              (None, 32, 32, 256)  590080      activation_12[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_13 (BatchNo (None, 32, 32, 256)  1024        conv2d_13[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "activation_13 (Activation)      (None, 32, 32, 256)  0           batch_normalization_13[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_transpose_2 (Conv2DTrans (None, 64, 64, 128)  131200      activation_13[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "concatenate_2 (Concatenate)     (None, 64, 64, 256)  0           conv2d_transpose_2[0][0]         \n",
            "                                                                 activation_3[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_14 (Conv2D)              (None, 64, 64, 128)  295040      concatenate_2[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_14 (BatchNo (None, 64, 64, 128)  512         conv2d_14[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "activation_14 (Activation)      (None, 64, 64, 128)  0           batch_normalization_14[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_15 (Conv2D)              (None, 64, 64, 128)  147584      activation_14[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_15 (BatchNo (None, 64, 64, 128)  512         conv2d_15[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "activation_15 (Activation)      (None, 64, 64, 128)  0           batch_normalization_15[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_transpose_3 (Conv2DTrans (None, 128, 128, 64) 32832       activation_15[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "concatenate_3 (Concatenate)     (None, 128, 128, 128 0           conv2d_transpose_3[0][0]         \n",
            "                                                                 activation_1[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_16 (Conv2D)              (None, 128, 128, 64) 73792       concatenate_3[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_16 (BatchNo (None, 128, 128, 64) 256         conv2d_16[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "activation_16 (Activation)      (None, 128, 128, 64) 0           batch_normalization_16[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_17 (Conv2D)              (None, 128, 128, 64) 36928       activation_16[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_17 (BatchNo (None, 128, 128, 64) 256         conv2d_17[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "activation_17 (Activation)      (None, 128, 128, 64) 0           batch_normalization_17[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_18 (Conv2D)              (None, 128, 128, 4)  260         activation_17[0][0]              \n",
            "==================================================================================================\n",
            "Total params: 31,054,340\n",
            "Trainable params: 31,042,564\n",
            "Non-trainable params: 11,776\n",
            "__________________________________________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "LmcYufmuaeUz",
        "outputId": "1e547c95-01a1-4ec2-a5da-b321bc959826"
      },
      "source": [
        "history = model.fit(X_train, y_train_cat, \n",
        "                    batch_size = 16, \n",
        "                    verbose=1, \n",
        "                    epochs=50, \n",
        "                    validation_data=(X_test, y_test_cat), \n",
        "                    shuffle=False)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/50\n",
            "60/60 [==============================] - 40s 259ms/step - loss: 0.3228 - accuracy: 0.9091 - val_loss: 27.7621 - val_accuracy: 0.0338\n",
            "Epoch 2/50\n",
            "60/60 [==============================] - 14s 234ms/step - loss: 0.1244 - accuracy: 0.9553 - val_loss: 0.8999 - val_accuracy: 0.7886\n",
            "Epoch 3/50\n",
            "60/60 [==============================] - 14s 240ms/step - loss: 0.1043 - accuracy: 0.9605 - val_loss: 1.3919 - val_accuracy: 0.7264\n",
            "Epoch 4/50\n",
            "60/60 [==============================] - 14s 240ms/step - loss: 0.0979 - accuracy: 0.9625 - val_loss: 1.0386 - val_accuracy: 0.7995\n",
            "Epoch 5/50\n",
            "60/60 [==============================] - 15s 246ms/step - loss: 0.0934 - accuracy: 0.9633 - val_loss: 1.3862 - val_accuracy: 0.7157\n",
            "Epoch 6/50\n",
            "60/60 [==============================] - 15s 248ms/step - loss: 0.0888 - accuracy: 0.9649 - val_loss: 1.2477 - val_accuracy: 0.7554\n",
            "Epoch 7/50\n",
            "60/60 [==============================] - 15s 251ms/step - loss: 0.0858 - accuracy: 0.9658 - val_loss: 1.3460 - val_accuracy: 0.7248\n",
            "Epoch 8/50\n",
            "60/60 [==============================] - 15s 252ms/step - loss: 0.0840 - accuracy: 0.9664 - val_loss: 1.0440 - val_accuracy: 0.7647\n",
            "Epoch 9/50\n",
            "60/60 [==============================] - 15s 258ms/step - loss: 0.0801 - accuracy: 0.9678 - val_loss: 0.9890 - val_accuracy: 0.7161\n",
            "Epoch 10/50\n",
            "60/60 [==============================] - 16s 264ms/step - loss: 0.0772 - accuracy: 0.9689 - val_loss: 0.5836 - val_accuracy: 0.7849\n",
            "Epoch 11/50\n",
            "60/60 [==============================] - 16s 265ms/step - loss: 0.0748 - accuracy: 0.9697 - val_loss: 0.3555 - val_accuracy: 0.8630\n",
            "Epoch 12/50\n",
            "60/60 [==============================] - 16s 261ms/step - loss: 0.0719 - accuracy: 0.9709 - val_loss: 0.2300 - val_accuracy: 0.9078\n",
            "Epoch 13/50\n",
            "60/60 [==============================] - 16s 259ms/step - loss: 0.0707 - accuracy: 0.9713 - val_loss: 0.1527 - val_accuracy: 0.9384\n",
            "Epoch 14/50\n",
            "60/60 [==============================] - 16s 260ms/step - loss: 0.0685 - accuracy: 0.9723 - val_loss: 0.2306 - val_accuracy: 0.9038\n",
            "Epoch 15/50\n",
            "60/60 [==============================] - 16s 262ms/step - loss: 0.0669 - accuracy: 0.9728 - val_loss: 0.2124 - val_accuracy: 0.9095\n",
            "Epoch 16/50\n",
            "60/60 [==============================] - 16s 263ms/step - loss: 0.0652 - accuracy: 0.9735 - val_loss: 0.1418 - val_accuracy: 0.9397\n",
            "Epoch 17/50\n",
            "60/60 [==============================] - 16s 263ms/step - loss: 0.0641 - accuracy: 0.9739 - val_loss: 0.1517 - val_accuracy: 0.9367\n",
            "Epoch 18/50\n",
            "60/60 [==============================] - 16s 262ms/step - loss: 0.0626 - accuracy: 0.9746 - val_loss: 0.1745 - val_accuracy: 0.9252\n",
            "Epoch 19/50\n",
            "60/60 [==============================] - 16s 263ms/step - loss: 0.0617 - accuracy: 0.9749 - val_loss: 0.1135 - val_accuracy: 0.9534\n",
            "Epoch 20/50\n",
            "60/60 [==============================] - 16s 263ms/step - loss: 0.0600 - accuracy: 0.9755 - val_loss: 0.1986 - val_accuracy: 0.9167\n",
            "Epoch 21/50\n",
            "60/60 [==============================] - 16s 263ms/step - loss: 0.0589 - accuracy: 0.9760 - val_loss: 0.1512 - val_accuracy: 0.9377\n",
            "Epoch 22/50\n",
            "60/60 [==============================] - 16s 261ms/step - loss: 0.0577 - accuracy: 0.9765 - val_loss: 0.1196 - val_accuracy: 0.9512\n",
            "Epoch 23/50\n",
            "60/60 [==============================] - 16s 261ms/step - loss: 0.0566 - accuracy: 0.9769 - val_loss: 0.1386 - val_accuracy: 0.9441\n",
            "Epoch 24/50\n",
            "60/60 [==============================] - 16s 262ms/step - loss: 0.0553 - accuracy: 0.9775 - val_loss: 0.1299 - val_accuracy: 0.9482\n",
            "Epoch 25/50\n",
            "60/60 [==============================] - 16s 266ms/step - loss: 0.0544 - accuracy: 0.9778 - val_loss: 0.1392 - val_accuracy: 0.9445\n",
            "Epoch 26/50\n",
            "60/60 [==============================] - 16s 264ms/step - loss: 0.0536 - accuracy: 0.9782 - val_loss: 0.2019 - val_accuracy: 0.9166\n",
            "Epoch 27/50\n",
            "60/60 [==============================] - 16s 265ms/step - loss: 0.0530 - accuracy: 0.9783 - val_loss: 0.1615 - val_accuracy: 0.9343\n",
            "Epoch 28/50\n",
            "60/60 [==============================] - 16s 263ms/step - loss: 0.0522 - accuracy: 0.9787 - val_loss: 0.1147 - val_accuracy: 0.9544\n",
            "Epoch 29/50\n",
            "60/60 [==============================] - 16s 265ms/step - loss: 0.0516 - accuracy: 0.9790 - val_loss: 0.1147 - val_accuracy: 0.9552\n",
            "Epoch 30/50\n",
            "60/60 [==============================] - 16s 264ms/step - loss: 0.0509 - accuracy: 0.9793 - val_loss: 0.1081 - val_accuracy: 0.9573\n",
            "Epoch 31/50\n",
            "60/60 [==============================] - 16s 264ms/step - loss: 0.0502 - accuracy: 0.9795 - val_loss: 0.0872 - val_accuracy: 0.9652\n",
            "Epoch 32/50\n",
            "60/60 [==============================] - 16s 262ms/step - loss: 0.0495 - accuracy: 0.9798 - val_loss: 0.0744 - val_accuracy: 0.9700\n",
            "Epoch 33/50\n",
            "60/60 [==============================] - 16s 263ms/step - loss: 0.0487 - accuracy: 0.9800 - val_loss: 0.1179 - val_accuracy: 0.9542\n",
            "Epoch 34/50\n",
            "60/60 [==============================] - 16s 262ms/step - loss: 0.0481 - accuracy: 0.9803 - val_loss: 0.1038 - val_accuracy: 0.9590\n",
            "Epoch 35/50\n",
            "60/60 [==============================] - 16s 264ms/step - loss: 0.0475 - accuracy: 0.9805 - val_loss: 0.0692 - val_accuracy: 0.9721\n",
            "Epoch 36/50\n",
            "60/60 [==============================] - 16s 261ms/step - loss: 0.0470 - accuracy: 0.9807 - val_loss: 0.0971 - val_accuracy: 0.9624\n",
            "Epoch 37/50\n",
            "60/60 [==============================] - 16s 264ms/step - loss: 0.0463 - accuracy: 0.9810 - val_loss: 0.0650 - val_accuracy: 0.9741\n",
            "Epoch 38/50\n",
            "60/60 [==============================] - 16s 262ms/step - loss: 0.0461 - accuracy: 0.9811 - val_loss: 0.0769 - val_accuracy: 0.9698\n",
            "Epoch 39/50\n",
            "60/60 [==============================] - 16s 262ms/step - loss: 0.0467 - accuracy: 0.9809 - val_loss: 0.0986 - val_accuracy: 0.9620\n",
            "Epoch 40/50\n",
            "60/60 [==============================] - 16s 263ms/step - loss: 0.0446 - accuracy: 0.9817 - val_loss: 0.0639 - val_accuracy: 0.9746\n",
            "Epoch 41/50\n",
            "60/60 [==============================] - 16s 261ms/step - loss: 0.0439 - accuracy: 0.9819 - val_loss: 0.0734 - val_accuracy: 0.9708\n",
            "Epoch 42/50\n",
            "60/60 [==============================] - 16s 261ms/step - loss: 0.0434 - accuracy: 0.9821 - val_loss: 0.0664 - val_accuracy: 0.9731\n",
            "Epoch 43/50\n",
            "60/60 [==============================] - 16s 262ms/step - loss: 0.0428 - accuracy: 0.9824 - val_loss: 0.0643 - val_accuracy: 0.9735\n",
            "Epoch 44/50\n",
            "60/60 [==============================] - 16s 263ms/step - loss: 0.0425 - accuracy: 0.9824 - val_loss: 0.0551 - val_accuracy: 0.9774\n",
            "Epoch 45/50\n",
            "60/60 [==============================] - 16s 265ms/step - loss: 0.0421 - accuracy: 0.9827 - val_loss: 0.0525 - val_accuracy: 0.9786\n",
            "Epoch 46/50\n",
            "60/60 [==============================] - 16s 265ms/step - loss: 0.0413 - accuracy: 0.9830 - val_loss: 0.0525 - val_accuracy: 0.9786\n",
            "Epoch 47/50\n",
            "60/60 [==============================] - 16s 263ms/step - loss: 0.0403 - accuracy: 0.9834 - val_loss: 0.0647 - val_accuracy: 0.9737\n",
            "Epoch 48/50\n",
            "60/60 [==============================] - 16s 262ms/step - loss: 0.0396 - accuracy: 0.9837 - val_loss: 0.1080 - val_accuracy: 0.9578\n",
            "Epoch 49/50\n",
            "60/60 [==============================] - 16s 261ms/step - loss: 0.0411 - accuracy: 0.9831 - val_loss: 0.1664 - val_accuracy: 0.9354\n",
            "Epoch 50/50\n",
            "60/60 [==============================] - 16s 260ms/step - loss: 0.0423 - accuracy: 0.9826 - val_loss: 0.0535 - val_accuracy: 0.9785\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KQ_NuTmlafmh"
      },
      "source": [
        "#Save the model for future use\n",
        "model.save('/content/drive/MyDrive/Colab Notebooks/saved_models/tutorial119_sandstone_50epochs.hdf5')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 573
        },
        "id": "iVibZ6c1ao7P",
        "outputId": "c9fb159a-53cd-456a-aff2-b0af62640283"
      },
      "source": [
        "###\n",
        "#plot the training and validation accuracy and loss at each epoch\n",
        "loss = history.history['loss']\n",
        "val_loss = history.history['val_loss']\n",
        "epochs = range(1, len(loss) + 1)\n",
        "plt.plot(epochs, loss, 'y', label='Training loss')\n",
        "plt.plot(epochs, val_loss, 'r', label='Validation loss')\n",
        "plt.title('Training and validation loss')\n",
        "plt.xlabel('Epochs')\n",
        "plt.ylabel('Loss')\n",
        "plt.legend()\n",
        "plt.show()\n",
        "\n",
        "acc = history.history['accuracy']\n",
        "val_acc = history.history['val_accuracy']\n",
        "\n",
        "plt.plot(epochs, acc, 'y', label='Training Accuracy')\n",
        "plt.plot(epochs, val_acc, 'r', label='Validation Accuracy')\n",
        "plt.title('Training and validation Accuracy')\n",
        "plt.xlabel('Epochs')\n",
        "plt.ylabel('Accuracy')\n",
        "plt.legend()\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RO7ACKoBapui"
      },
      "source": [
        "\n",
        "        #Load previously saved model\n",
        "from keras.models import load_model\n",
        "model = load_model(\"/content/drive/MyDrive/Colab Notebooks/saved_models/tutorial119_sandstone_50epochs.hdf5\", compile=False)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "UNuftNH18ZoY"
      },
      "source": [
        "y_pred=model.predict(X_test)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "b_SF6IU38cxw",
        "outputId": "b1a4ac40-a3e9-40e7-95d7-b249ed13f6c6"
      },
      "source": [
        "y_pred.shape"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(40, 128, 128, 4)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 18
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "drpgFz-SjPq3",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "6934927b-6610-4784-b8dd-f75c7c320a78"
      },
      "source": [
        "y_pred_argmax=np.argmax(y_pred, axis=3)\n",
        "y_pred_argmax.shape"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(40, 128, 128)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 19
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "5panJLTGawJx",
        "outputId": "d195ca02-d42d-4388-a587-4ce8d7a7fcb7"
      },
      "source": [
        "#Using built in keras function\n",
        "from keras.metrics import MeanIoU\n",
        "n_classes = 4\n",
        "IOU_keras = MeanIoU(num_classes=n_classes)  \n",
        "IOU_keras.update_state(y_test[:,:,:,0], y_pred_argmax)\n",
        "print(\"Mean IoU =\", IOU_keras.result().numpy())"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Mean IoU = 0.92037\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "hgHHa8q9axAx",
        "outputId": "de3de859-7ca6-4f81-8d2d-a97501f6f3a1"
      },
      "source": [
        "#To calculate I0U for each class...\n",
        "values = np.array(IOU_keras.get_weights()).reshape(n_classes, n_classes)\n",
        "print(values)\n",
        "class1_IoU = values[0,0]/(values[0,0] + values[0,1] + values[0,2] + values[0,3] + values[1,0]+ values[2,0]+ values[3,0])\n",
        "class2_IoU = values[1,1]/(values[1,1] + values[1,0] + values[1,2] + values[1,3] + values[0,1]+ values[2,1]+ values[3,1])\n",
        "class3_IoU = values[2,2]/(values[2,2] + values[2,0] + values[2,1] + values[2,3] + values[0,2]+ values[1,2]+ values[3,2])\n",
        "class4_IoU = values[3,3]/(values[3,3] + values[3,0] + values[3,1] + values[3,2] + values[0,3]+ values[1,3]+ values[2,3])\n",
        "\n",
        "print(\"IoU for class1 is: \", class1_IoU)\n",
        "print(\"IoU for class2 is: \", class2_IoU)\n",
        "print(\"IoU for class3 is: \", class3_IoU)\n",
        "print(\"IoU for class4 is: \", class4_IoU)\n",
        "\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[[1.03736e+05 7.37000e+02 4.82000e+02 0.00000e+00]\n",
            " [3.53700e+03 3.60980e+04 1.87300e+03 0.00000e+00]\n",
            " [2.12300e+03 2.98700e+03 4.87613e+05 4.37000e+02]\n",
            " [1.00000e+00 0.00000e+00 1.78000e+02 1.55580e+04]]\n",
            "IoU for class1 is:  0.93780285\n",
            "IoU for class2 is:  0.79806334\n",
            "IoU for class3 is:  0.98369956\n",
            "IoU for class4 is:  0.9619142\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 261
        },
        "id": "zW-IdHfAa2f6",
        "outputId": "68a6eb40-1f74-4f90-b489-6375a473151f"
      },
      "source": [
        "#Predict on a few images\n",
        "#model = get_model()\n",
        "#model.load_weights('???.hdf5')  \n",
        "import random\n",
        "test_img_number = random.randint(0, len(X_test)-1)\n",
        "test_img = X_test[test_img_number]\n",
        "ground_truth=y_test[test_img_number]\n",
        "test_img_norm=test_img[:,:,0][:,:,None]\n",
        "test_img_input=np.expand_dims(test_img_norm, 0)\n",
        "prediction = (model.predict(test_img_input))\n",
        "predicted_img=np.argmax(prediction, axis=3)[0,:,:]\n",
        "\n",
        "\n",
        "plt.figure(figsize=(12, 8))\n",
        "plt.subplot(231)\n",
        "plt.title('Testing Image')\n",
        "plt.imshow(test_img[:,:,0], cmap='gray')\n",
        "plt.subplot(232)\n",
        "plt.title('Testing Label')\n",
        "plt.imshow(ground_truth[:,:,0], cmap='jet')\n",
        "plt.subplot(233)\n",
        "plt.title('Prediction on test image')\n",
        "plt.imshow(predicted_img, cmap='jet')\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 864x576 with 3 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    }
  ]
}