task_2.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "20d46d11-a1f4-4d25-b5ee-b8102f7bca27",
   "metadata": {},
   "source": [
    "## Task2: OCR and Sentiment Analysis"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "638c539b-1013-4507-bfe5-4e230d3a0f8a",
   "metadata": {},
   "source": [
    "#### Short explanations are given as comments wherever required"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fb652d45-2d39-4459-92b7-c1ed3da90899",
   "metadata": {},
   "source": [
    "#### Please run each cell to get complete final output"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3a48f416-6f70-4bf7-883d-1324be67aa96",
   "metadata": {},
   "source": [
    "### OCR "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "11734248-af75-43dd-9fdb-b6a614db04ee",
   "metadata": {},
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
    "from tensorflow.keras.optimizers import RMSprop"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fb4ad626-52db-4295-aeea-29a176597ce7",
   "metadata": {},
   "source": [
    "##### Cell 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "821a2502-1e9f-44d7-beb6-a0eb4f5e8546",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Converting the csv data to image\n",
    "\n",
    "# sent=pd.read_csv('alphabets_28x28.csv')\n",
    "# sent=sent.dropna()\n",
    "# sent=sent.drop_duplicates()\n",
    "# sent=sent.reset_index(drop=True)\n",
    "# g=[]\n",
    "# for i in range(sent['label'].nunique()):\n",
    "#     sum=0\n",
    "#     for j in range(i+1):\n",
    "#         sum+=sent['label'].value_counts().sort_index().iloc[j]\n",
    "#     g.append(sum-1)    \n",
    "# l=[0]   \n",
    "# for i in range(len(g)-1):\n",
    "#     l.append(g[i]+1)\n",
    "# name_1=np.zeros((26*500,1),dtype='str')\n",
    "# label_1=np.zeros((26*500,1),dtype='str')\n",
    "# name=[]\n",
    "# label=[]\n",
    "# for i in range(1,(26*500)+1):\n",
    "#     name.append(f'grayscale_image_{i}.png')  \n",
    "#     if (i%500)!=0:\n",
    "#         label.append((sent['label'].unique())[i//500])\n",
    "#     else:\n",
    "#         label.append((sent['label'].unique())[(i//500)-1])\n",
    "# data=(\n",
    "#     name,\n",
    "#     label\n",
    "# )\n",
    "# excel=pd.DataFrame(data,index=[\"Name\",\"Label\"])\n",
    "# width, height = 28,28\n",
    "# for k in range(26):\n",
    "#     count=0\n",
    "#     for z in range(l[k],g[k]):\n",
    "#         if count<500:\n",
    "#             grayscale_image = np.zeros((width,height))\n",
    "#             for i in range(width):\n",
    "#                 for j in range(height):\n",
    "#                     h=(i*width)+j+1\n",
    "#                     grayscale_image[i][j]=sent.iloc[z][h]\n",
    "#             grayscale_image=(grayscale_image).astype('uint8')\n",
    "#             # Convert the numpy array to a PIL image\n",
    "#             image = Image.fromarray(grayscale_image)\n",
    "#             count+=1\n",
    "#             #image.show()\n",
    "#             image.save(f'grayscale_image_{(k*500)+count}.png')\n",
    "            "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "570ec59e-880f-480a-ad56-0cf1edc2f7eb",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 10400 images belonging to 26 classes.\n",
      "Found 2600 images belonging to 26 classes.\n"
     ]
    }
   ],
   "source": [
    "# Spliting the dataset into training,validation dataset\n",
    "output_dir= 'alphabet_28x28_images'\n",
    "\n",
    "img_height, img_width = 28, 28\n",
    "batch_size = 32\n",
    "\n",
    "train_datagen = ImageDataGenerator(\n",
    "    rescale=1./255,        \n",
    "    validation_split=0.2   \n",
    ")\n",
    "\n",
    "\n",
    "train_generator = train_datagen.flow_from_directory(\n",
    "    output_dir,\n",
    "    target_size=(img_height, img_width),\n",
    "    batch_size=batch_size,\n",
    "    color_mode='grayscale',\n",
    "    class_mode='categorical',\n",
    "    subset='training'\n",
    ")\n",
    "\n",
    "validation_generator = train_datagen.flow_from_directory(\n",
    "    output_dir,\n",
    "    target_size=(img_height, img_width),\n",
    "    batch_size=batch_size,\n",
    "    color_mode='grayscale',\n",
    "    class_mode='categorical',\n",
    "    subset='validation'\n",
    ")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1a5dec1a-753a-487e-baee-3224d8f24c44",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\vishn\\AppData\\Local\\Programs\\Python\\Python312\\Lib\\site-packages\\keras\\src\\layers\\convolutional\\base_conv.py:107: UserWarning: Do not pass an `input_shape`/`input_dim` argument to a layer. When using Sequential models, prefer using an `Input(shape)` object as the first layer in the model instead.\n",
      "  super().__init__(activity_regularizer=activity_regularizer, **kwargs)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/20\n",
      "\u001b[1m325/325\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m16s\u001b[0m 43ms/step - acc: 0.0358 - loss: 0.3737 - val_acc: 0.0765 - val_loss: 0.1649\n",
      "Epoch 2/20\n",
      "\u001b[1m325/325\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m10s\u001b[0m 31ms/step - acc: 0.0744 - loss: 0.1638 - val_acc: 0.0650 - val_loss: 0.1616\n",
      "Epoch 3/20\n",
      "\u001b[1m325/325\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m10s\u001b[0m 30ms/step - acc: 0.1415 - loss: 0.1603 - val_acc: 0.2654 - val_loss: 0.1553\n",
      "Epoch 4/20\n",
      "\u001b[1m325/325\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m10s\u001b[0m 30ms/step - acc: 0.3102 - loss: 0.1499 - val_acc: 0.3758 - val_loss: 0.1294\n",
      "Epoch 5/20\n",
      "\u001b[1m325/325\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m10s\u001b[0m 29ms/step - acc: 0.4796 - loss: 0.1180 - val_acc: 0.5531 - val_loss: 0.0987\n",
      "Epoch 6/20\n",
      "\u001b[1m325/325\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m10s\u001b[0m 29ms/step - acc: 0.6288 - loss: 0.0869 - val_acc: 0.6296 - val_loss: 0.0806\n",
      "Epoch 7/20\n",
      "\u001b[1m325/325\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m10s\u001b[0m 30ms/step - acc: 0.7092 - loss: 0.0677 - val_acc: 0.6827 - val_loss: 0.0700\n",
      "Epoch 8/20\n",
      "\u001b[1m325/325\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m10s\u001b[0m 30ms/step - acc: 0.7528 - loss: 0.0586 - val_acc: 0.7204 - val_loss: 0.0628\n",
      "Epoch 9/20\n",
      "\u001b[1m325/325\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m10s\u001b[0m 30ms/step - acc: 0.7988 - loss: 0.0500 - val_acc: 0.7435 - val_loss: 0.0575\n",
      "Epoch 10/20\n",
      "\u001b[1m325/325\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m9s\u001b[0m 29ms/step - acc: 0.8184 - loss: 0.0450 - val_acc: 0.7627 - val_loss: 0.0525\n",
      "Epoch 11/20\n",
      "\u001b[1m325/325\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m10s\u001b[0m 29ms/step - acc: 0.8385 - loss: 0.0410 - val_acc: 0.7896 - val_loss: 0.0497\n",
      "Epoch 12/20\n",
      "\u001b[1m325/325\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m10s\u001b[0m 30ms/step - acc: 0.8572 - loss: 0.0371 - val_acc: 0.8038 - val_loss: 0.0473\n",
      "Epoch 13/20\n",
      "\u001b[1m325/325\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m10s\u001b[0m 30ms/step - acc: 0.8713 - loss: 0.0343 - val_acc: 0.8123 - val_loss: 0.0447\n",
      "Epoch 14/20\n",
      "\u001b[1m325/325\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m10s\u001b[0m 29ms/step - acc: 0.8773 - loss: 0.0320 - val_acc: 0.8231 - val_loss: 0.0432\n",
      "Epoch 15/20\n",
      "\u001b[1m 98/325\u001b[0m \u001b[32m━━━━━━\u001b[0m\u001b[37m━━━━━━━━━━━━━━\u001b[0m \u001b[1m5s\u001b[0m 24ms/step - acc: 0.8759 - loss: 0.0315"
     ]
    }
   ],
   "source": [
    "# Building and Training of the CNN Model\n",
    "mod = tf.keras.models.Sequential([\n",
    "    tf.keras.layers.Conv2D(32, (3, 3), activation='relu', input_shape=(img_height, img_width, 1)),\n",
    "    tf.keras.layers.MaxPooling2D(2, 2),\n",
    "    tf.keras.layers.Conv2D(64, (3, 3), activation='relu'),\n",
    "    tf.keras.layers.MaxPooling2D(2, 2),\n",
    "    tf.keras.layers.Conv2D(128, (3, 3), activation='relu'),\n",
    "    tf.keras.layers.MaxPooling2D(2, 2),\n",
    "    tf.keras.layers.Flatten(),\n",
    "    tf.keras.layers.Dense(512, activation='relu'),\n",
    "    tf.keras.layers.Dense(len(train_generator.class_indices), activation='softmax')\n",
    "])\n",
    "\n",
    "\n",
    "mod.compile(optimizer = tf.keras.optimizers.RMSprop(learning_rate=0.0001), loss = 'binary_crossentropy',metrics = ['acc'])\n",
    "\n",
    "\n",
    "history = mod.fit(\n",
    "    train_generator,\n",
    "    validation_data=validation_generator,\n",
    "    epochs=20\n",
    ")\n",
    "mod.save('tensorflow_ocr.h5')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "98f39a2a-f890-4c53-aa09-31f574c48012",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:absl:Compiled the loaded model, but the compiled metrics have yet to be built. `model.compile_metrics` will be empty until you train or evaluate the model.\n"
     ]
    }
   ],
   "source": [
    "mod=tf.keras.models.load_model('tensorflow_ocr.h5')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "67cc802c-6df0-465d-a55d-f7a2eee8fd02",
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import cv2\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd\n",
    "from PIL import Image"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "0be785ee-148e-4be2-8348-90094a587c5e",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\vishn\\AppData\\Local\\Temp\\ipykernel_22064\\2243735585.py:2: DtypeWarning: Columns (1,2,3,4) have mixed types. Specify dtype option on import or set low_memory=False.\n",
      "  sent=pd.read_csv('alphabets_28x28.csv')\n"
     ]
    }
   ],
   "source": [
    "# Loading the csv file\n",
    "sent=pd.read_csv('alphabets_28x28.csv')\n",
    "sent=sent.dropna()\n",
    "sent=sent.drop_duplicates()\n",
    "sent=sent.reset_index(drop=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "18211e7b-c77b-48d0-b77f-bcc145755a99",
   "metadata": {},
   "source": [
    "##### Cell 8"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "63b4810a-f861-44ae-b8ba-fb2145f478f3",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 540ms/step\n",
      "This image is probably S\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 47ms/step\n",
      "This image is probably D\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 42ms/step\n",
      "This image is probably A\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 47ms/step\n",
      "This image is probably J\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 32ms/step\n",
      "This image is probably I\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 32ms/step\n",
      "This image is probably B\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 32ms/step\n",
      "This image is probably F\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 32ms/step\n",
      "This image is probably K\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 48ms/step\n",
      "This image is probably K\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 48ms/step\n",
      "This image is probably K\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 47ms/step\n",
      "This image is probably R\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 49ms/step\n",
      "This image is probably I\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Testing the model with custom dataset\n",
    "c=1\n",
    "while os.path.isfile(f\"test_images/image_{c}.png\"):\n",
    "    img=cv2.imread(f\"test_images/image_{c}.png\")[:,:,0]\n",
    "    img=np.invert(np.array([img]))\n",
    "    pred=mod.predict(img)\n",
    "    print(f\"This image is probably {sent['label'].unique()[np.argmax(pred)]}\")\n",
    "    plt.imshow(img[0],cmap=plt.cm.binary)\n",
    "    plt.show()\n",
    "    c+=1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "34f1e10a-dc10-43bf-9da3-f86bbe50d9c5",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:absl:Compiled the loaded model, but the compiled metrics have yet to be built. `model.compile_metrics` will be empty until you train or evaluate the model.\n"
     ]
    }
   ],
   "source": [
    "# Loading the saved model\n",
    "# Image Segmentation and drawing contours\n",
    "\n",
    "mod=tf.keras.models.load_model('tensorflow_ocr_1.h5')\n",
    "def segment_image(image_path):\n",
    "    #  image in grayscale\n",
    "    image = cv2.imread(image_path, cv2.IMREAD_GRAYSCALE)\n",
    "    \n",
    "    \n",
    "    _, binary_image = cv2.threshold(image, 128, 255, cv2.THRESH_BINARY | cv2.THRESH_OTSU)\n",
    "    \n",
    "    contours, _ = cv2.findContours(binary_image, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)\n",
    "    \n",
    "    \n",
    "    segmented_image = cv2.cvtColor(image, cv2.COLOR_GRAY2BGR)\n",
    "    cv2.drawContours(segmented_image, contours, -1, (0, 255, 0), 2)\n",
    "    \n",
    "    return image, contours\n",
    "\n",
    "image_path = 'target_images/line_2.png'  \n",
    "original_image, contours = segment_image(image_path)\n",
    "\n",
    "segmented_image = cv2.cvtColor(original_image, cv2.COLOR_GRAY2BGR)\n",
    "cv2.drawContours(segmented_image, contours, -1, (0, 255, 0), 2)\n",
    "cv2.imshow('Segmented Image', segmented_image)\n",
    "cv2.waitKey(0)\n",
    "cv2.destroyAllWindows()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "51b257ce-3777-48ee-9245-92cccec8351f",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Preprocessing the segmented image\n",
    "def pad_image_to_28x28(image):\n",
    "    h, w = image.shape[:2]\n",
    "    top = (28 - h) // 2\n",
    "    bottom = 28 - h - top\n",
    "    left = (28 - w) // 2\n",
    "    right = 28 - w - left\n",
    "    padded_image = cv2.copyMakeBorder(image, top, bottom, left, right, cv2.BORDER_CONSTANT, value=0)\n",
    "    return padded_image\n",
    "def preprocess_region(region):\n",
    "    # Resize the region to match the input size of the OCR model (e.g., 28x28)\n",
    "    region = pad_image_to_28x28(region)\n",
    "    region = region.astype('float32') / 255.0  # Normalize the region\n",
    "    region = np.expand_dims(region, axis=-1)  # Add channel dimension\n",
    "    region = np.expand_dims(region, axis=0)   # Add batch dimension\n",
    "    return region\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c0cc055b-f5a3-48dd-beb0-4fd113735ab7",
   "metadata": {},
   "outputs": [],
   "source": [
    "def recognize_characters_with_whitespace(image, contours, model, whitespace_threshold=10):\n",
    "    contours = sorted(contours, key=lambda c: cv2.boundingRect(c)[0])\n",
    "    recognized_text = \"\"\n",
    "    previous_x = None\n",
    "    for contour in contours:\n",
    "        x, y, w, h = cv2.boundingRect(contour)\n",
    "        region = image[y:y+h, x:x+w]\n",
    "        processed_region = preprocess_region(region)\n",
    "        prediction = model.predict(processed_region)\n",
    "        character = chr(np.argmax(prediction) + ord('A'))  # Adjust according to your model's output\n",
    "        if previous_x is not None and (x - previous_x) > whitespace_threshold:\n",
    "            recognized_text += \" \"\n",
    "        recognized_text += character\n",
    "        previous_x = x + w\n",
    "    return recognized_text\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "510ebfdf-a44b-4523-8dec-7501837a81df",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 129ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 18ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 35ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 18ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 18ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 34ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 20ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 32ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 31ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 33ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 31ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 22ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 30ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 31ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 31ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 33ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 31ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 31ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 31ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 22ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 21ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 16ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 31ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 32ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 36ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 43ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 44ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 22ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 22ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 25ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 22ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 18ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 19ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 18ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 18ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 18ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 16ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 16ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 21ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 17ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 17ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 18ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 31ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 20ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 20ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 16ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 19ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 20ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 18ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 23ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 20ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 32ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 44ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 44ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 28ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 19ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 24ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 31ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 29ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 32ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 45ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 42ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 43ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 45ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 38ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 21ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 38ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 43ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 40ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 37ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 43ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 29ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 35ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 46ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 38ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 37ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 23ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 18ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 31ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 47ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 33ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 33ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 32ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 34ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 31ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 38ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 31ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 31ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 31ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 31ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 31ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 37ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 38ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 29ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 24ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 31ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 32ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 42ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 30ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 39ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 31ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 31ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 31ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 28ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 40ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 31ms/step\n",
      "Recognized Text: SDUREPRGOAONEGITCFFAYLLPERSUISCOFFOKCALRUOYHODMASISJGUINESDIRUNEDDITNTTAYAPRENTVEUOYATTAHGNIRTAAATSSURFHII\n"
     ]
    }
   ],
   "source": [
    "# OCR of obtained text from the processed image using the model\n",
    "def recognize_characters(image, contours, model):\n",
    "    recognized_text = \"\"\n",
    "    for contour in contours:\n",
    "        x, y, w, h = cv2.boundingRect(contour)\n",
    "        region = image[y:y+h, x:x+w]\n",
    "        processed_region = preprocess_region(region)\n",
    "        prediction = model.predict(processed_region)\n",
    "        character = sent['label'].unique()[np.argmax(prediction)]  \n",
    "        recognized_text += character\n",
    "    return recognized_text\n",
    "\n",
    "\n",
    "recognized_text = recognize_characters_with_whitespace(original_image, contours, ocr_model)\n",
    "print('Recognized Text:', recognized_text)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e6bc93c8-ba0f-440a-a637-2a6aab2d9b15",
   "metadata": {},
   "source": [
    "### Sentiment Analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "74cd1b4b-de0d-4121-abe2-67d3afc72a8f",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Loading sentiment source file\n",
    "info=pd.read_csv('sentiment_analysis_dataset.csv')\n",
    "info_1 =pd.read_csv('sentiment_analysis_dataset.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "22391321-a865-44ae-b2f5-c8e324a25e7f",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Finding probability of each sentiment(in the order of info['sentiment'].unique())\n",
    "p_sent=[]\n",
    "for i in range(info['sentiment'].nunique()):\n",
    "    p_sent.append((info['sentiment'].value_counts().sort_index().iloc[i])/info['sentiment'].count())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "90232ef2-7ffc-4267-98f0-7bd64c8ebc82",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Finding word count in each document\n",
    "word_count=[]\n",
    "for i in range(info['line'].count()):\n",
    "    count=0\n",
    "    for index, char in enumerate(info['line'].iloc[i]):\n",
    "        if char.isspace():\n",
    "            count+=1\n",
    "    word_count.append(count+1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "8323b5c2-831e-490b-bd19-7b7d882e4fed",
   "metadata": {},
   "outputs": [],
   "source": [
    "word=word_count"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "f78f026c-003b-4fac-82a9-ca8b850451fc",
   "metadata": {},
   "outputs": [],
   "source": [
    "#converting info['line'] to str datatype\n",
    "info['line']=info['line'].astype('str')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "fcceb4fe-35b4-41be-92a9-a48fc96cbbf6",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Spilting each string to identify different words present in them\n",
    "# Adding all words to vocab list\n",
    "info['line']=info['line'].str.split()\n",
    "vocab=[]\n",
    "l=info['line'].iloc[0]\n",
    "vocab.append(l[0])\n",
    "for i in range(info['line'].count()):\n",
    "    l=info['line'].iloc[i]\n",
    "    for j in range(word_count[i]):\n",
    "        if l[j] not in vocab:\n",
    "            vocab.append(l[j])       "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "efb62542-c042-4001-a72d-d235c3ea73e3",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Finding count of each vocab word wrt all given sentiments\n",
    "final=np.zeros(shape=(info['sentiment'].nunique(),len(vocab)),dtype='int')\n",
    "for i in range(info['line'].count()):\n",
    "    tot_1=np.zeros(shape=(info['sentiment'].nunique(),len(vocab)),dtype='int')\n",
    "    l=info['line'].iloc[i]  \n",
    "    g=np.where(info['sentiment'].unique()==info['sentiment'].iloc[i])[0]\n",
    "    for j in range(word[i]):\n",
    "        tot_1[g[0]][vocab.index(l[j])]+=1\n",
    "    for j in range(len(vocab)):\n",
    "        if tot_1[g[0]][j]>1:\n",
    "            tot_1[g[0]][j]=1    \n",
    "    final+=tot_1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "64d5e3e7-e350-417e-a3cd-0074be6da2a5",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Parameter estimation using the obtained count\n",
    "para=np.zeros(shape=(info['sentiment'].nunique(),len(vocab)))\n",
    "sum=[]\n",
    "for i in range(info['sentiment'].nunique()):\n",
    "    sm=0\n",
    "    for j in range(len(vocab)):\n",
    "        sm+=final[i][j]\n",
    "    sum.append(sm+len(vocab))\n",
    "for i in range(info['sentiment'].nunique()):\n",
    "    for j in range(len(vocab)):\n",
    "        para[i][j]=(final[i][j]+1)/sum[i]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "98868640-99a8-418a-a22d-ac0d789c45c5",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Implementation of Sentiment Analysis\n",
    "target=\n",
    "cn=0\n",
    "for i in range(len(target)):\n",
    "    if target[i].isspace():\n",
    "        cn+=1   \n",
    "cn+=1\n",
    "li=[]\n",
    "kim=0\n",
    "for i in range(cn):\n",
    "    gu=\"\"\n",
    "    for j in range(kim,len(target)):\n",
    "        if target[j].isspace():\n",
    "            kim=j+1\n",
    "            break\n",
    "        else:\n",
    "            gu+=target[j]\n",
    "    li.append(gu)\n",
    "test_data=[]\n",
    "for i in range(len(li)):\n",
    "    if li[i] in vocab:\n",
    "        test_data.append(li[i])\n",
    "test_data      \n",
    "test_index=[]\n",
    "for i in range(len(test_data)):\n",
    "    test_index.append(vocab.index(test_data[i]))    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "81b9957e-9227-4421-b128-5ee983c06f4b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Given data is probably on the ' Angry ' note\n"
     ]
    }
   ],
   "source": [
    "# Finding the most probable sentiment\n",
    "max=0\n",
    "best=0\n",
    "prob=[]\n",
    "for i in range(info_1['sentiment'].nunique()):\n",
    "    p=1\n",
    "    for j in range(len(test_data)):\n",
    "        p*=para[i][test_index[j]]\n",
    "    p*=p_sent[i]\n",
    "    prob.append(p)\n",
    "    if i==0:\n",
    "        max=p\n",
    "        best=i\n",
    "    else:\n",
    "        if max<p:\n",
    "            max=p\n",
    "            best=i            \n",
    "print(\"Given data is probably on the '\",info['sentiment'].unique()[best],\"' note\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "df0844dc-7b0a-40dd-ae7c-1e996da187d7",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.12.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}