main.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "from keras.preprocessing.image import ImageDataGenerator\n",
    "import pandas as pd\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Conv2D, MaxPooling2D, Flatten, Dense\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# untuk test\n",
    "from keras.preprocessing.image import img_to_array\n",
    "import cv2\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataset_path = 'dataset-img'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Baca file CSV\n",
    "df = pd.read_csv('data-label/hiragana_labels.csv')\n",
    "\n",
    "# Konversi DataFrame ke dictionary\n",
    "hiragana_labels = dict(zip(df.Hiragana, zip(df.Romaji, df.Label)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'あ': ('a', 0), 'い': ('i', 1), 'う': ('u', 2), 'え': ('e', 3), 'お': ('o', 4), 'か': ('ka', 5), 'き': ('ki', 6), 'く': ('ku', 7), 'け': ('ke', 8), 'こ': ('ko', 9), 'さ': ('sa', 10), 'し': ('shi', 11), 'す': ('su', 12), 'せ': ('se', 13), 'そ': ('so', 14), 'た': ('ta', 15), 'ち': ('chi', 16), 'つ': ('tsu', 17), 'て': ('te', 18), 'と': ('to', 19), 'な': ('na', 20), 'に': ('ni', 21), 'ぬ': ('nu', 22), 'ね': ('ne', 23), 'の': ('no', 24), 'は': ('ha', 25), 'ひ': ('hi', 26), 'ふ': ('fu', 27), 'へ': ('he', 28), 'ほ': ('ho', 29), 'ま': ('ma', 30), 'み': ('mi', 31), 'む': ('mu', 32), 'め': ('me', 33), 'も': ('mo', 34), 'や': ('ya', 35), 'ゆ': ('yu', 36), 'よ': ('yo', 37), 'ら': ('ra', 38), 'り': ('ri', 39), 'る': ('ru', 40), 'れ': ('re', 41), 'ろ': ('ro', 42), 'わ': ('wa', 43), 'を': ('wo', 44), 'ん': ('n', 45)}\n"
     ]
    }
   ],
   "source": [
    "print(hiragana_labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 768 images belonging to 56 classes.\n",
      "Found 192 images belonging to 56 classes.\n"
     ]
    }
   ],
   "source": [
    "# Fungsi untuk mengonversi label ke bentuk numerik\n",
    "def label_to_numeric(label):\n",
    "    return hiragana_labels[label]\n",
    "\n",
    "# Gunakan ImageDataGenerator untuk memuat dan melabeli gambar\n",
    "datagen = ImageDataGenerator(rescale=1.0/255.0, validation_split=0.2)\n",
    "\n",
    "train_generator = datagen.flow_from_directory(\n",
    "    dataset_path,\n",
    "    target_size=(32, 32),  # sesuaikan dengan ukuran input model\n",
    "    color_mode='grayscale',  # sesuaikan dengan warna gambar\n",
    "    batch_size=32,\n",
    "    class_mode='categorical',\n",
    "    subset='training'\n",
    ")\n",
    "\n",
    "validation_generator = datagen.flow_from_directory(\n",
    "    dataset_path,\n",
    "    target_size=(32, 32),\n",
    "    color_mode='grayscale',\n",
    "    batch_size=32,\n",
    "    class_mode='categorical',\n",
    "    subset='validation'\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Label: れ, Romaji: re\n",
      "Label: None, Romaji: None\n",
      "Label: け, Romaji: ke\n",
      "Label: ね, Romaji: ne\n",
      "Label: は, Romaji: ha\n",
      "Label: つ, Romaji: tsu\n",
      "Label: は, Romaji: ha\n",
      "Label: う, Romaji: u\n",
      "Label: None, Romaji: None\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\blptn\\AppData\\Local\\Temp\\ipykernel_7932\\1042434735.py:18: UserWarning: Glyph 12428 (\\N{HIRAGANA LETTER RE}) missing from current font.\n",
      "  plt.tight_layout()\n",
      "C:\\Users\\blptn\\AppData\\Local\\Temp\\ipykernel_7932\\1042434735.py:18: UserWarning: Glyph 12369 (\\N{HIRAGANA LETTER KE}) missing from current font.\n",
      "  plt.tight_layout()\n",
      "C:\\Users\\blptn\\AppData\\Local\\Temp\\ipykernel_7932\\1042434735.py:18: UserWarning: Glyph 12397 (\\N{HIRAGANA LETTER NE}) missing from current font.\n",
      "  plt.tight_layout()\n",
      "C:\\Users\\blptn\\AppData\\Local\\Temp\\ipykernel_7932\\1042434735.py:18: UserWarning: Glyph 12399 (\\N{HIRAGANA LETTER HA}) missing from current font.\n",
      "  plt.tight_layout()\n",
      "C:\\Users\\blptn\\AppData\\Local\\Temp\\ipykernel_7932\\1042434735.py:18: UserWarning: Glyph 12388 (\\N{HIRAGANA LETTER TU}) missing from current font.\n",
      "  plt.tight_layout()\n",
      "C:\\Users\\blptn\\AppData\\Local\\Temp\\ipykernel_7932\\1042434735.py:18: UserWarning: Glyph 12358 (\\N{HIRAGANA LETTER U}) missing from current font.\n",
      "  plt.tight_layout()\n",
      "C:\\Users\\blptn\\AppData\\Roaming\\Python\\Python39\\site-packages\\IPython\\core\\pylabtools.py:152: UserWarning: Glyph 12428 (\\N{HIRAGANA LETTER RE}) missing from current font.\n",
      "  fig.canvas.print_figure(bytes_io, **kw)\n",
      "C:\\Users\\blptn\\AppData\\Roaming\\Python\\Python39\\site-packages\\IPython\\core\\pylabtools.py:152: UserWarning: Glyph 12369 (\\N{HIRAGANA LETTER KE}) missing from current font.\n",
      "  fig.canvas.print_figure(bytes_io, **kw)\n",
      "C:\\Users\\blptn\\AppData\\Roaming\\Python\\Python39\\site-packages\\IPython\\core\\pylabtools.py:152: UserWarning: Glyph 12397 (\\N{HIRAGANA LETTER NE}) missing from current font.\n",
      "  fig.canvas.print_figure(bytes_io, **kw)\n",
      "C:\\Users\\blptn\\AppData\\Roaming\\Python\\Python39\\site-packages\\IPython\\core\\pylabtools.py:152: UserWarning: Glyph 12388 (\\N{HIRAGANA LETTER TU}) missing from current font.\n",
      "  fig.canvas.print_figure(bytes_io, **kw)\n",
      "C:\\Users\\blptn\\AppData\\Roaming\\Python\\Python39\\site-packages\\IPython\\core\\pylabtools.py:152: UserWarning: Glyph 12358 (\\N{HIRAGANA LETTER U}) missing from current font.\n",
      "  fig.canvas.print_figure(bytes_io, **kw)\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x1000 with 9 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Fungsi untuk mengonversi label one-hot encoded ke label sesuai dengan kamus hiragana_labels\n",
    "def one_hot_to_label(one_hot_label):\n",
    "    for label, (romaji, class_id) in hiragana_labels.items():\n",
    "        if one_hot_label[class_id] == 1:\n",
    "            return label, romaji\n",
    "    return None, None  # Return None jika label tidak ditemukan\n",
    "\n",
    "# Tampilkan beberapa contoh gambar dari dataset pelatihan\n",
    "train_sample_images, train_sample_labels = next(train_generator)\n",
    "plt.figure(figsize=(10, 10))\n",
    "for i in range(9):\n",
    "    plt.subplot(3, 3, i + 1)\n",
    "    plt.imshow(train_sample_images[i].reshape(32, 32), cmap='gray')\n",
    "    label, romaji = one_hot_to_label(train_sample_labels[i])\n",
    "    print(f\"Label: {label}, Romaji: {romaji}\")  # Tampilkan label dan romaji\n",
    "    plt.title(f'Training Label: {label} ({romaji})')\n",
    "    plt.axis('off')\n",
    "plt.tight_layout()\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = Sequential([\n",
    "    Conv2D(32, kernel_size=(3, 3), activation='relu', input_shape=(32, 32, 1)),\n",
    "    MaxPooling2D(pool_size=(2, 2)),\n",
    "    Conv2D(64, kernel_size=(3, 3), activation='relu'),\n",
    "    MaxPooling2D(pool_size=(2, 2)),\n",
    "    Flatten(),\n",
    "    Dense(120, activation='relu'),\n",
    "    Dense(84, activation='relu'),\n",
    "    Dense(train_generator.num_classes, activation='softmax')  # Sesuaikan dengan jumlah kelas\n",
    "])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/10\n",
      "24/24 [==============================] - 3s 64ms/step - loss: 3.8633 - accuracy: 0.0573 - val_loss: 3.4390 - val_accuracy: 0.0885\n",
      "Epoch 2/10\n",
      "24/24 [==============================] - 1s 37ms/step - loss: 2.4901 - accuracy: 0.3294 - val_loss: 1.3204 - val_accuracy: 0.6146\n",
      "Epoch 3/10\n",
      "24/24 [==============================] - 1s 40ms/step - loss: 0.8984 - accuracy: 0.7292 - val_loss: 0.6949 - val_accuracy: 0.7917\n",
      "Epoch 4/10\n",
      "24/24 [==============================] - 1s 48ms/step - loss: 0.5331 - accuracy: 0.8451 - val_loss: 0.5085 - val_accuracy: 0.8906\n",
      "Epoch 5/10\n",
      "24/24 [==============================] - 1s 35ms/step - loss: 0.2555 - accuracy: 0.9154 - val_loss: 0.4085 - val_accuracy: 0.8750\n",
      "Epoch 6/10\n",
      "24/24 [==============================] - 1s 37ms/step - loss: 0.1317 - accuracy: 0.9688 - val_loss: 0.4670 - val_accuracy: 0.8542\n",
      "Epoch 7/10\n",
      "24/24 [==============================] - 1s 36ms/step - loss: 0.0647 - accuracy: 0.9831 - val_loss: 0.3041 - val_accuracy: 0.9062\n",
      "Epoch 8/10\n",
      "24/24 [==============================] - 1s 38ms/step - loss: 0.0306 - accuracy: 0.9961 - val_loss: 0.3348 - val_accuracy: 0.9115\n",
      "Epoch 9/10\n",
      "24/24 [==============================] - 1s 44ms/step - loss: 0.0378 - accuracy: 0.9870 - val_loss: 0.4377 - val_accuracy: 0.8698\n",
      "Epoch 10/10\n",
      "24/24 [==============================] - 1s 37ms/step - loss: 0.0326 - accuracy: 0.9896 - val_loss: 0.3685 - val_accuracy: 0.8750\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.src.callbacks.History at 0x1e389486e20>"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# tambahan model\n",
    "model.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy'])\n",
    "\n",
    "# Latih model\n",
    "model.fit(\n",
    "    train_generator,\n",
    "    epochs=10,  # Sesuaikan dengan kebutuhan\n",
    "    validation_data=validation_generator\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "6/6 [==============================] - 0s 22ms/step - loss: 0.3685 - accuracy: 0.8750\n",
      "Akurasi pada data validasi: 87.50%\n"
     ]
    }
   ],
   "source": [
    "# Evaluasi model\n",
    "test_loss, test_acc = model.evaluate(validation_generator)\n",
    "print(f'Akurasi pada data validasi: {test_acc * 100:.2f}%')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\Users\\blptn\\AppData\\Local\\Programs\\Python\\Python39\\lib\\site-packages\\keras\\src\\engine\\training.py:3000: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
      "  saving_api.save_model(\n"
     ]
    }
   ],
   "source": [
    "from keras.models import load_model\n",
    "\n",
    "# Definisikan path lengkap untuk penyimpanan model\n",
    "model_path = 'model/model.h5'  # Ganti dengan path lengkap Anda\n",
    "\n",
    "# Simpan model ke file model.h5\n",
    "model.save(model_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "def predict_image_cv(model, image, label_map):\n",
    "    # Ubah gambar menjadi skala abu-abu\n",
    "    gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)\n",
    "    # Resize gambar ke ukuran yang diharapkan oleh model\n",
    "    resized_img = cv2.resize(gray, (32, 32))\n",
    "    # Normalisasi nilai pixel\n",
    "    resized_img = resized_img / 255.0\n",
    "    # Ubah gambar menjadi array numpy dan tambahkan dimensi batch\n",
    "    img_array = np.expand_dims(img_to_array(resized_img), axis=0)\n",
    "    img_array = np.expand_dims(img_array, axis=-1)\n",
    "\n",
    "    # Melakukan prediksi menggunakan model\n",
    "    predictions = model.predict(img_array)\n",
    "\n",
    "    # Mendapatkan kelas prediksi\n",
    "    predicted_class = np.argmax(predictions[0])\n",
    "\n",
    "    # Mendapatkan label dari kelas prediksi\n",
    "    for label, (romaji, class_id) in label_map.items():\n",
    "        if class_id == predicted_class:\n",
    "            predicted_label = label\n",
    "            predicted_romaji = romaji\n",
    "            break\n",
    "\n",
    "    return predicted_label, predicted_romaji"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 119ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 18ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 44ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 36ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 18ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 39ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 18ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 18ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 17ms/step\n",
      "1/1 [==============================] - 0s 18ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 29ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 29ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 139ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 18ms/step\n",
      "1/1 [==============================] - 0s 18ms/step\n",
      "1/1 [==============================] - 0s 18ms/step\n",
      "1/1 [==============================] - 0s 18ms/step\n"
     ]
    }
   ],
   "source": [
    "def predict_image_cv(model, image, label_map):\n",
    "    # Ubah gambar menjadi skala abu-abu\n",
    "    gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)\n",
    "    # Resize gambar ke ukuran yang diharapkan oleh model\n",
    "    resized_img = cv2.resize(gray, (32, 32))\n",
    "    # Normalisasi nilai pixel\n",
    "    resized_img = resized_img / 255.0\n",
    "    # Ubah gambar menjadi array numpy dan tambahkan dimensi batch\n",
    "    img_array = np.expand_dims(img_to_array(resized_img), axis=0)\n",
    "    img_array = np.expand_dims(img_array, axis=-1)\n",
    "\n",
    "    # Melakukan prediksi menggunakan model\n",
    "    predictions = model.predict(img_array)\n",
    "\n",
    "    # Mendapatkan kelas prediksi\n",
    "    predicted_class = np.argmax(predictions[0])\n",
    "\n",
    "    # Mendapatkan label dari kelas prediksi\n",
    "    predicted_label = None\n",
    "    predicted_romaji = None\n",
    "    for label, (romaji, class_id) in label_map.items():\n",
    "        if class_id == predicted_class:\n",
    "            predicted_label = label\n",
    "            predicted_romaji = romaji\n",
    "            break\n",
    "\n",
    "    return predicted_label, predicted_romaji\n",
    "\n",
    "# Inisialisasi kamera\n",
    "cap = cv2.VideoCapture(0)\n",
    "\n",
    "# Loop untuk mendapatkan frame dari kamera\n",
    "while True:\n",
    "    # Baca frame dari kamera\n",
    "    ret, frame = cap.read()\n",
    "    if not ret:\n",
    "        print(\"Gagal membaca frame dari kamera.\")\n",
    "        break\n",
    "\n",
    "    # Lakukan prediksi pada frame yang didapat\n",
    "    predicted_label, predicted_romaji = predict_image_cv(model, frame, hiragana_labels)\n",
    "\n",
    "    # Tampilkan hasil prediksi pada layar\n",
    "    if predicted_label is not None and predicted_romaji is not None:\n",
    "        cv2.putText(frame, f'Prediksi: {predicted_label} ({predicted_romaji})', (10, 30), cv2.FONT_HERSHEY_SIMPLEX, 0.8, (0, 255, 0), 2)\n",
    "    cv2.imshow('Frame', frame)\n",
    "\n",
    "    # Keluar dari loop jika tombol 'q' ditekan\n",
    "    if cv2.waitKey(1) & 0xFF == ord('q'):\n",
    "        break\n",
    "\n",
    "# Tutup kamera dan jendela OpenCV\n",
    "cap.release()\n",
    "cv2.destroyAllWindows()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "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": "3.9.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}