Image_Processing_MNIST.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Image_Processing_MNIST.ipynb",
      "provenance": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "l8T3s5Ft1Fkf"
      },
      "source": [
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "import keras\n",
        "from keras.datasets import mnist\n",
        "from keras.layers import Dense\n",
        "from keras.optimizers import Adam\n",
        "from keras.models import Sequential\n",
        "from keras.utils.np_utils import to_categorical\n",
        "import random"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "UKdLJ2OH2kvC"
      },
      "source": [
        "np.random.seed(0)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "t_0dDgr9tMCE"
      },
      "source": [
        "(X_train, y_train), (X_test, y_test) = mnist.load_data()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "D2OJuiTrtbZn",
        "outputId": "6634aa6e-d034-45e5-906b-179e27b6b18c",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "print(X_train.shape)\n",
        "print(X_test.shape)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(60000, 28, 28)\n",
            "(10000, 28, 28)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "WYXYUpnG2EXW"
      },
      "source": [
        "assert(X_train.shape[0] == y_train.shape[0]), \"The number of images is equal to the number of labels\"\n",
        "assert(X_test.shape[0] == y_test.shape[0]), \"The number of images is equal to the number of labels\"\n",
        "assert(X_train.shape[1:] == (28, 28)), \"The dimensions of the images are not 28x28\"\n",
        "assert(X_test.shape[1:] == (28, 28)), \"The dimensions of the images are not 28x28\""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SReDrSSM21vM",
        "outputId": "89d4020a-664f-47a5-9491-37dfdc85029d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 718
        }
      },
      "source": [
        "num_of_samples = []\n",
        "\n",
        "cols = 5\n",
        "num_classes = 10\n",
        "\n",
        "fig, axs = plt.subplots(nrows=num_classes, ncols=cols, figsize=(5,10))\n",
        "fig.tight_layout()\n",
        "for i in range(cols):\n",
        "  for j in range(num_classes):\n",
        "    x_selected = X_train[y_train == j]\n",
        "    axs[j][i].imshow(x_selected[random.randint(0, len(x_selected - 1)), :, :], cmap=plt.get_cmap(\"gray\"))\n",
        "    axs[j][i].axis(\"off\")\n",
        "    if i == 2:\n",
        "      axs[j][i].set_title(str(j))\n",
        "      num_of_samples.append(len(x_selected))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 360x720 with 50 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vkHiL5CUKsOb",
        "outputId": "586a9c9a-7cf1-4c50-f7e0-e02674ef8e08",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 329
        }
      },
      "source": [
        "print(num_of_samples)\n",
        "plt.figure(figsize=(12,4))\n",
        "plt.bar(range(0, num_classes), num_of_samples)\n",
        "plt.title(\"Distribution of the Training Dataset\")\n",
        "plt.xlabel(\"Class Number\")\n",
        "plt.ylabel(\"Number of Images\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[5923, 6742, 5958, 6131, 5842, 5421, 5918, 6265, 5851, 5949]\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0, 0.5, 'Number of Images')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 149
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 864x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZFXbJ3uwN4p3"
      },
      "source": [
        "y_train = to_categorical(y_train, 10)\n",
        "y_test = to_categorical(y_test, 10)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TRnToE9d72fx"
      },
      "source": [
        "X_train = X_train/255\n",
        "X_test = X_test/255"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9W-V5RR-OK5r",
        "outputId": "e3f73e9b-6c59-459a-9e37-8fdcca3f4cef",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "num_pixels = 784\n",
        "X_train = X_train.reshape(X_train.shape[0], num_pixels)\n",
        "X_test = X_test.reshape(X_test.shape[0], num_pixels)\n",
        "print(X_train.shape)\n",
        "print(X_test.shape)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(60000, 784)\n",
            "(10000, 784)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MP-eH7nVQNeX"
      },
      "source": [
        "def create_model():\n",
        "  model = Sequential()\n",
        "  model.add(Dense(10, input_dim=num_pixels, activation='relu'))\n",
        "  model.add(Dense(30, activation='relu'))\n",
        "  model.add(Dense(10, activation='relu'))\n",
        "  model.add(Dense(num_classes, activation='softmax'))\n",
        "  model.compile(Adam(learning_rate=0.01), loss='categorical_crossentropy', metrics=['accuracy'])\n",
        "  return model"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "daWhqeiDRsui",
        "outputId": "9147ccea-27e2-45a3-865d-2a97a0752cb2",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "model = create_model()\n",
        "print(model.summary())"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential_7\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "dense_22 (Dense)             (None, 10)                7850      \n",
            "_________________________________________________________________\n",
            "dense_23 (Dense)             (None, 30)                330       \n",
            "_________________________________________________________________\n",
            "dense_24 (Dense)             (None, 10)                310       \n",
            "_________________________________________________________________\n",
            "dense_25 (Dense)             (None, 10)                110       \n",
            "=================================================================\n",
            "Total params: 8,600\n",
            "Trainable params: 8,600\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n",
            "None\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bQRAYNrnTm2P",
        "outputId": "0d7cc777-1553-47f7-c98a-ce878cf774ed",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "history = model.fit(x=X_train, y=y_train, validation_split=0.1, epochs=30, verbose=1, shuffle=1)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/30\n",
            "1688/1688 [==============================] - 3s 1ms/step - loss: 0.4963 - accuracy: 0.8508 - val_loss: 0.3268 - val_accuracy: 0.9082\n",
            "Epoch 2/30\n",
            "1688/1688 [==============================] - 2s 1ms/step - loss: 0.3807 - accuracy: 0.8909 - val_loss: 0.3038 - val_accuracy: 0.9082\n",
            "Epoch 3/30\n",
            "1688/1688 [==============================] - 2s 1ms/step - loss: 0.3516 - accuracy: 0.8997 - val_loss: 0.2813 - val_accuracy: 0.9192\n",
            "Epoch 4/30\n",
            "1688/1688 [==============================] - 2s 1ms/step - loss: 0.3328 - accuracy: 0.9056 - val_loss: 0.2895 - val_accuracy: 0.9193\n",
            "Epoch 5/30\n",
            "1688/1688 [==============================] - 3s 2ms/step - loss: 0.3231 - accuracy: 0.9084 - val_loss: 0.2683 - val_accuracy: 0.9222\n",
            "Epoch 6/30\n",
            "1688/1688 [==============================] - 3s 2ms/step - loss: 0.3134 - accuracy: 0.9123 - val_loss: 0.2837 - val_accuracy: 0.9200\n",
            "Epoch 7/30\n",
            "1688/1688 [==============================] - 3s 2ms/step - loss: 0.3077 - accuracy: 0.9135 - val_loss: 0.2714 - val_accuracy: 0.9242\n",
            "Epoch 8/30\n",
            "1688/1688 [==============================] - 3s 2ms/step - loss: 0.3006 - accuracy: 0.9160 - val_loss: 0.2776 - val_accuracy: 0.9225\n",
            "Epoch 9/30\n",
            "1688/1688 [==============================] - 2s 1ms/step - loss: 0.2971 - accuracy: 0.9166 - val_loss: 0.2508 - val_accuracy: 0.9233\n",
            "Epoch 10/30\n",
            "1688/1688 [==============================] - 2s 1ms/step - loss: 0.2943 - accuracy: 0.9184 - val_loss: 0.2811 - val_accuracy: 0.9225\n",
            "Epoch 11/30\n",
            "1688/1688 [==============================] - 2s 1ms/step - loss: 0.2841 - accuracy: 0.9217 - val_loss: 0.2666 - val_accuracy: 0.9215\n",
            "Epoch 12/30\n",
            "1688/1688 [==============================] - 2s 1ms/step - loss: 0.2928 - accuracy: 0.9188 - val_loss: 0.2763 - val_accuracy: 0.9215\n",
            "Epoch 13/30\n",
            "1688/1688 [==============================] - 2s 1ms/step - loss: 0.2860 - accuracy: 0.9219 - val_loss: 0.2587 - val_accuracy: 0.9317\n",
            "Epoch 14/30\n",
            "1688/1688 [==============================] - 2s 1ms/step - loss: 0.2811 - accuracy: 0.9230 - val_loss: 0.2635 - val_accuracy: 0.9227\n",
            "Epoch 15/30\n",
            "1688/1688 [==============================] - 2s 1ms/step - loss: 0.2822 - accuracy: 0.9229 - val_loss: 0.2528 - val_accuracy: 0.9338\n",
            "Epoch 16/30\n",
            "1688/1688 [==============================] - 2s 1ms/step - loss: 0.2823 - accuracy: 0.9221 - val_loss: 0.2633 - val_accuracy: 0.9302\n",
            "Epoch 17/30\n",
            "1688/1688 [==============================] - 2s 1ms/step - loss: 0.2754 - accuracy: 0.9239 - val_loss: 0.2596 - val_accuracy: 0.9307\n",
            "Epoch 18/30\n",
            "1688/1688 [==============================] - 2s 1ms/step - loss: 0.2767 - accuracy: 0.9235 - val_loss: 0.2452 - val_accuracy: 0.9303\n",
            "Epoch 19/30\n",
            "1688/1688 [==============================] - 2s 1ms/step - loss: 0.2721 - accuracy: 0.9249 - val_loss: 0.2443 - val_accuracy: 0.9328\n",
            "Epoch 20/30\n",
            "1688/1688 [==============================] - 2s 1ms/step - loss: 0.2723 - accuracy: 0.9264 - val_loss: 0.2669 - val_accuracy: 0.9290\n",
            "Epoch 21/30\n",
            "1688/1688 [==============================] - 2s 1ms/step - loss: 0.2746 - accuracy: 0.9244 - val_loss: 0.2569 - val_accuracy: 0.9307\n",
            "Epoch 22/30\n",
            "1688/1688 [==============================] - 2s 1ms/step - loss: 0.2737 - accuracy: 0.9256 - val_loss: 0.2773 - val_accuracy: 0.9287\n",
            "Epoch 23/30\n",
            "1688/1688 [==============================] - 2s 1ms/step - loss: 0.2726 - accuracy: 0.9260 - val_loss: 0.2869 - val_accuracy: 0.9220\n",
            "Epoch 24/30\n",
            "1688/1688 [==============================] - 2s 1ms/step - loss: 0.2755 - accuracy: 0.9251 - val_loss: 0.2974 - val_accuracy: 0.9235\n",
            "Epoch 25/30\n",
            "1688/1688 [==============================] - 2s 1ms/step - loss: 0.2714 - accuracy: 0.9263 - val_loss: 0.2498 - val_accuracy: 0.9328\n",
            "Epoch 26/30\n",
            "1688/1688 [==============================] - 2s 1ms/step - loss: 0.2689 - accuracy: 0.9276 - val_loss: 0.2664 - val_accuracy: 0.9277\n",
            "Epoch 27/30\n",
            "1688/1688 [==============================] - 2s 1ms/step - loss: 0.2745 - accuracy: 0.9243 - val_loss: 0.2947 - val_accuracy: 0.9248\n",
            "Epoch 28/30\n",
            "1688/1688 [==============================] - 2s 1ms/step - loss: 0.2684 - accuracy: 0.9257 - val_loss: 0.2761 - val_accuracy: 0.9312\n",
            "Epoch 29/30\n",
            "1688/1688 [==============================] - 2s 1ms/step - loss: 0.2663 - accuracy: 0.9269 - val_loss: 0.2630 - val_accuracy: 0.9312\n",
            "Epoch 30/30\n",
            "1688/1688 [==============================] - 2s 1ms/step - loss: 0.2703 - accuracy: 0.9263 - val_loss: 0.2544 - val_accuracy: 0.9305\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hpd3CzqsUE8f",
        "outputId": "4455de9f-0478-4ef4-834f-d76de42da0a7",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "plt.plot(history.history['loss'])\n",
        "plt.plot(history.history['val_loss'])\n",
        "plt.legend(['loss', 'validation_loss'])\n",
        "plt.title('Loss')\n",
        "plt.xlabel('epoch')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0.5, 0, 'epoch')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 156
        },
        {
          "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": "UAk7TJMsYTzG",
        "outputId": "816f18e1-e35e-4f11-98d3-43c56a4d9445",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "plt.plot(history.history['accuracy'])\n",
        "plt.plot(history.history['val_accuracy'])\n",
        "plt.legend(['accuracy', 'validation_accuracy'])\n",
        "plt.title('accuracy')\n",
        "plt.xlabel('epoch')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0.5, 0, 'epoch')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 157
        },
        {
          "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": "JpWPrzEEduBj",
        "outputId": "77b0a7cf-784b-4c3e-a074-1cfeb57e2303",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "score = model.evaluate(x=X_test, y=y_test, verbose=0)\n",
        "print(type(score))\n",
        "print('Test score:', score[0])\n",
        "print('Test accuracy:', score[1])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "<class 'list'>\n",
            "Test score: 0.29959118366241455\n",
            "Test accuracy: 0.921999990940094\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zCyKk6xV1Fau",
        "outputId": "206af11e-632b-412a-9048-02b0dd0b2411",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "import requests\n",
        "from PIL import Image\n",
        "url = 'https://colah.github.io/posts/2014-10-Visualizing-MNIST/img/mnist_pca/MNIST-p1815-4.png'\n",
        "response = requests.get(url, stream=True)\n",
        "img = Image.open(response.raw)\n",
        "plt.imshow(img)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.image.AxesImage at 0x7f5ee804e438>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 159
        },
        {
          "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": "SZQyAXiM8RMA",
        "outputId": "969e4a54-124f-4e27-a6eb-e044c3c91fa5",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "import cv2\n",
        "img_array = np.asarray(img)\n",
        "resized = cv2.resize(img_array, (28, 28))\n",
        "gray_scale = cv2.cvtColor(resized, cv2.COLOR_BGR2GRAY)\n",
        "image = cv2.bitwise_not(gray_scale)\n",
        "plt.imshow(image, cmap=plt.get_cmap(\"gray\"))\n",
        "print(image)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[[  0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0\n",
            "    0   0   0   0   0   0   0   0   0   0]\n",
            " [  0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0\n",
            "    0   0   0   0   0   0   0   0   0 210]\n",
            " [  0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0\n",
            "    0   0   0   0   0   0   0   0   0 210]\n",
            " [  0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0\n",
            "    0   0   0   0   0   0   0   0   0 210]\n",
            " [  0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0  67 170\n",
            "  254 155 100  17   0   0   0   0   0 210]\n",
            " [  0   0   0   0   0   0   0   0   0   0   0  18  89 176 155 198 239 252\n",
            "  252 252 252 174   0   0   0   0   0 209]\n",
            " [  0   0   0   0   0   0   0   0   0   0  27 234 251 252 142  28  28  28\n",
            "   62 222 252 246  87   0   0   0   0 210]\n",
            " [  0   0   0   0   0   0   0   0   0   0  28 228 243 229 119   2   2   2\n",
            "   18 185 252 247  88   0   0   0   0 210]\n",
            " [  0   0   0   0   0   0   0   0   0   0   0   0  69   0   0   0   0   0\n",
            "    0 161 252 198   0   0   0   0   0 210]\n",
            " [  0   0   0   0   0   0   0   0   0   0   0   0   4   0   0   0   0   0\n",
            "  104 242 252 198   0   0   0   0   0 210]\n",
            " [  0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0\n",
            "  215 252 252  86   0   0   0   0   0 210]\n",
            " [  0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0  60\n",
            "  234 252 196  21   0   0   0   0   0 210]\n",
            " [  0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0  61 240\n",
            "  179 250 100   0   0   0   0   0   0 210]\n",
            " [  0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   6 176 252\n",
            "  252 244   0   0   0   0   0   0   0 210]\n",
            " [  0   0   0   0   0   0   0   0   0  71 123 123 123 123 123 158 252 252\n",
            "  252 169   0   0   0   0   0   0   0 210]\n",
            " [  0   0   0   0   0   0   0  54 221 239 243 182 246 252 252 252 252 252\n",
            "  197  10   0   0   0   0   0   0   0 210]\n",
            " [  0   0   0   0   0   0 115 187 251 212  73   0 136 252 252 252 252 252\n",
            "  190   0   0   0   0   0   0   0   0 210]\n",
            " [  0   0   0   0   0  27 233 246 140  58  14 115 240 252 252 252 252 252\n",
            "  218  41   0   0   0   0   0   0   0 210]\n",
            " [  0   0   0   0  28 196 234 176  44 143 221 252 252 251 213 180 223 252\n",
            "  252 223   0   0   0   0   0   0   0 210]\n",
            " [  0   0   0   0 100 252 191 150 208 236 252 252 250 230  46   0  70 248\n",
            "  252 244   4   0   0   0   0   0   0 210]\n",
            " [  0   0   0   0  91 235 252 252 252 239 222 167  66   0   0   0   0 115\n",
            "  252 252 216  29   0   0   0   0   0 210]\n",
            " [  0   0   0   0  72 193 238 238 238 205 163  40   3   0   0   0   0 115\n",
            "  252 252 246  42   0   0   0   0   0 210]\n",
            " [  0   0   0   0   0   5  23  23  23  10   0   0   0   0   0   0   0  27\n",
            "   59  59  43   4   0   0   0   0   0 209]\n",
            " [  0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0\n",
            "    0   0   0   0   0   0   0   0   0 210]\n",
            " [  0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0\n",
            "    0   0   0   0   0   0   0   0   0 210]\n",
            " [  0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0\n",
            "    0   0   0   0   0   0   0   0   0 210]\n",
            " [  0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0\n",
            "    0   0   0   0   0   0   0   0   0 210]\n",
            " [  0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0\n",
            "    0   0   0   0   0   0   0   0   0   0]]\n"
          ],
          "name": "stdout"
        },
        {
          "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": "LPoQZXC_87XF"
      },
      "source": [
        "image = image/255\n",
        "image = image.reshape(1, 784)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "72PY58pH9GTf",
        "outputId": "6ca2c0a7-8b29-4b6d-ef86-644045945988",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "prediction = model.predict_classes(image)\n",
        "print(\"Predicted image is\", str(prediction))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Predicted image is [2]\n"
          ],
          "name": "stdout"
        }
      ]
    }
  ]
}