how_i_made_.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# CAPTCHA Letter Extraction and Classification\n",
    "\n",
    "This notebook demonstrates the process of extracting letters from CAPTCHA images and classifying them using a deep learning model. The steps involved are:\n",
    "\n",
    "1. **Importing Libraries**: Importing necessary libraries for image processing, machine learning, and deep learning.\n",
    "2. **Helper Functions**: Defining helper functions for image display and letter extraction.\n",
    "3. **Data Preparation**: Loading CAPTCHA images, extracting letters, and preparing the dataset.\n",
    "4. **Data Augmentation**: Applying data augmentation techniques to increase the size and variability of the dataset.\n",
    "5. **Model Creation**: Creating and compiling a Convolutional Neural Network (CNN) model for letter classification.\n",
    "6. **Model Training**: Training the model with the prepared dataset and applying class weights to handle class imbalance.\n",
    "7. **Evaluation and Visualization**: Evaluating the model's performance and visualizing the results.\n",
    "8. **Testing on New CAPTCHAs**: Testing the trained model on new CAPTCHA images and displaying the results.\n",
    "\n",
    "Let's get started!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "from ultralytics import YOLO\n",
    "import matplotlib.pyplot as plt\n",
    "import os\n",
    "import numpy as np\n",
    "import cv2\n",
    "import tensorflow as tf\n",
    "from PIL import Image\n",
    "from tensorflow.keras.preprocessing.image import img_to_array\n",
    "from IPython.display import clear_output\n",
    "from sklearn.preprocessing import LabelEncoder, OneHotEncoder\n",
    "from sklearn.model_selection import train_test_split\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Activation, MaxPooling2D, Flatten, Conv2D, Dropout, Dense\n",
    "from tensorflow.keras.layers import Conv2D, MaxPooling2D, Flatten, Dense, Dropout, BatchNormalization, Activation\n",
    "from tensorflow.keras.models import Sequential\n",
    "from tensorflow.keras.callbacks import EarlyStopping, ReduceLROnPlateau\n",
    "from tensorflow.keras.models import save_model\n",
    "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
    "import time\n",
    "from tensorflow.keras.models import load_model\n",
    "from tensorflow.keras.utils import to_categorical\n",
    "import joblib\n",
    "from time import sleep"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "def show_image(image):\n",
    "    plt.close()\n",
    "    plt.imshow(image , cmap='gray')\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Letter Extraction from CAPTCHA Images\n",
    "\n",
    "In this section, we define a function to extract individual letters from CAPTCHA images using a pre-trained YOLO model. The function processes each image, detects the bounding boxes around the letters, and extracts and resizes each letter for further processing.\n",
    "\n",
    "### Function: `extract_letters_from_captcha`\n",
    "\n",
    "- **Parameters**:\n",
    "    - `image_path` (str): The path to the CAPTCHA image.\n",
    "    - `model_yolo` (YOLO): The pre-trained YOLO model for letter detection.\n",
    "\n",
    "- **Returns**:\n",
    "    - `letters` (list): A list of extracted letter images.\n",
    "    - `label` (list): The corresponding label for each letter.\n",
    "    - `error` (list): A list of errors encountered during processing.\n",
    "\n",
    "The function also includes a helper function `check_letter_tarakom` to verify the quality of the extracted letters.\n",
    "\n",
    "### Usage\n",
    "\n",
    "The function is used to process all CAPTCHA images in the dataset, extract the letters, and prepare them for training a deep learning model for letter classification."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "def extract_letters_from_captcha(image_path, model_yolo):\n",
    "    def check_letter_tarakom(pic):\n",
    "        _, pic = cv2.threshold(pic, 127, 255, cv2.THRESH_BINARY)\n",
    "        s = 90 - (np.sum(pic , axis=0 , keepdims=True) / 255)\n",
    "        total = len(s[0])\n",
    "        howmanyblack = 0\n",
    "        for i in s[0]:\n",
    "            if np.sum(i) >= 175:\n",
    "                howmanyblack += 1\n",
    "        if total - howmanyblack <= 22:\n",
    "            return True\n",
    "        else:\n",
    "            return False\n",
    "    letters = []\n",
    "    detected = []\n",
    "\n",
    "    how_many = 1\n",
    "\n",
    "    try:\n",
    "        image = cv2.imread(image_path)  \n",
    "\n",
    "\n",
    "        results = model_yolo(image_path)\n",
    "        detections = results[0].boxes\n",
    "        detections = sorted(detections, key=lambda box: box.xyxy[0][0]) \n",
    "        \n",
    "        last_one = []\n",
    "\n",
    "        howmany = 0 \n",
    "        for box in detections:\n",
    "            x1, y1, x2, y2 = map(int, box.xyxy[0].tolist())\n",
    "            if last_one != []:\n",
    "                if int(x1) - last_one[0] > 10:\n",
    "                    letter_crop = image[y1:y2, x1:x2] \n",
    "                    resized_letter = cv2.resize(letter_crop, (32, 52)) \n",
    "                    resized_letter = cv2.cvtColor(resized_letter, cv2.COLOR_BGR2GRAY) \n",
    "                    if howmany == 8:\n",
    "                        if check_letter_tarakom(letter_crop) == True:\n",
    "                            _, binarized = cv2.threshold(resized_letter, 128, 255, cv2.THRESH_BINARY)\n",
    "                            letters.append(resized_letter)\n",
    "                    elif howmany <= 7:\n",
    "                        _, binarized = cv2.threshold(resized_letter, 128, 255, cv2.THRESH_BINARY)\n",
    "                        letters.append(resized_letter)\n",
    "                    last_one = [int(x1), int(y1), int(x2), int(y2)]\n",
    "                    howmany += 1\n",
    "            else:\n",
    "                letter_crop = image[y1:y2, x1:x2] \n",
    "                resized_letter = cv2.resize(letter_crop, (32, 52)) \n",
    "                resized_letter = cv2.cvtColor(resized_letter, cv2.COLOR_BGR2GRAY)\n",
    "                if check_letter_tarakom(letter_crop) == True:\n",
    "                        _, binarized = cv2.threshold(resized_letter, 128, 255, cv2.THRESH_BINARY)\n",
    "                        letters.append(resized_letter)\n",
    "                last_one = [int(x1), int(y1), int(x2), int(y2)]\n",
    "                howmany += 1\n",
    "        print(len(letters))\n",
    "        error = []\n",
    "        if len(letters) == 5 or len(letters) == 8:\n",
    "            print(\"yes\")\n",
    "            label = os.path.basename(image_path.split(sep=\"/\")[1]).split(\".\")[0]\n",
    "            return letters, list(label) , error\n",
    "        else:\n",
    "            print(how_many , \" Ignored\")\n",
    "            how_many += 1\n",
    "            error.append(label)\n",
    "    except Exception as e:\n",
    "        print(f\"Error processing {image_path}: {e}\")\n",
    "        return [], []\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "files = os.listdir(\"kaggledataset\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'model_yolo' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "Cell \u001b[1;32mIn[14], line 4\u001b[0m\n\u001b[0;32m      2\u001b[0m labels \u001b[38;5;241m=\u001b[39m []\n\u001b[0;32m      3\u001b[0m \u001b[38;5;28;01mfor\u001b[39;00m file \u001b[38;5;129;01min\u001b[39;00m files:\n\u001b[1;32m----> 4\u001b[0m     result \u001b[38;5;241m=\u001b[39m extract_letters_from_captcha(\u001b[38;5;124mf\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mkaggledataset/\u001b[39m\u001b[38;5;132;01m{\u001b[39;00mfile\u001b[38;5;132;01m}\u001b[39;00m\u001b[38;5;124m\"\u001b[39m , model_yolo\u001b[38;5;241m=\u001b[39m\u001b[43mmodel_yolo\u001b[49m)\n\u001b[0;32m      5\u001b[0m     \u001b[38;5;28;01mtry\u001b[39;00m:\n\u001b[0;32m      6\u001b[0m         \u001b[38;5;28;01mfor\u001b[39;00m i \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28mrange\u001b[39m(\u001b[38;5;241m0\u001b[39m,\u001b[38;5;28mlen\u001b[39m(result[\u001b[38;5;241m0\u001b[39m])):\n",
      "\u001b[1;31mNameError\u001b[0m: name 'model_yolo' is not defined"
     ]
    }
   ],
   "source": [
    "images = []\n",
    "labels = []\n",
    "for file in files:\n",
    "    result = extract_letters_from_captcha(f\"kaggledataset/{file}\" , model_yolo=model_yolo)\n",
    "    try:\n",
    "        for i in range(0,len(result[0])):\n",
    "            images.append(result[0][i])\n",
    "        for i in result[1]:\n",
    "            labels.append(i)\n",
    "    except Exception as e:\n",
    "        print(\"-------------\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "images = []\n",
    "labels = []\n",
    "for file in files:\n",
    "    result = extract_letters_from_captcha(f\"kaggledataset/{file}\" , model_yolo=model_yolo)\n",
    "    try:\n",
    "        for i in range(0,len(result[0])):\n",
    "            images.append(result[0][i])\n",
    "        for i in result[1]:\n",
    "            labels.append(i)\n",
    "    except Exception as e:\n",
    "        print(\"-------------\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Min pixel value: 0.0\n",
      "Max pixel value: 0.87058824\n"
     ]
    }
   ],
   "source": [
    "images = np.array(images)\n",
    "labels = np.array(labels)\n",
    "images = images.astype('float32') / 255.0 "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "ename": "KeyboardInterrupt",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
      "Cell \u001b[1;32mIn[16], line 35\u001b[0m\n\u001b[0;32m     32\u001b[0m img \u001b[38;5;241m=\u001b[39m np\u001b[38;5;241m.\u001b[39mexpand_dims(img, axis\u001b[38;5;241m=\u001b[39m\u001b[38;5;241m0\u001b[39m)  \u001b[38;5;66;03m# اضافه کردن بعد batch\u001b[39;00m\n\u001b[0;32m     34\u001b[0m \u001b[38;5;28;01mfor\u001b[39;00m _ \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28mrange\u001b[39m(\u001b[38;5;241m20\u001b[39m):  \u001b[38;5;66;03m# تولید 10 تصویر Augmented برای هر تصویر\u001b[39;00m\n\u001b[1;32m---> 35\u001b[0m     aug_img \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mnext\u001b[39;49m\u001b[43m(\u001b[49m\u001b[43mdatagen\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mflow\u001b[49m\u001b[43m(\u001b[49m\u001b[43mimg\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mbatch_size\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;241;43m1\u001b[39;49m\u001b[43m)\u001b[49m\u001b[43m)\u001b[49m[\u001b[38;5;241m0\u001b[39m]\u001b[38;5;241m.\u001b[39mastype(np\u001b[38;5;241m.\u001b[39muint8)\n\u001b[0;32m     36\u001b[0m     aug_img \u001b[38;5;241m=\u001b[39m aug_img\u001b[38;5;241m.\u001b[39msqueeze()  \u001b[38;5;66;03m# حذف ابعاد اضافی\u001b[39;00m\n\u001b[0;32m     37\u001b[0m     aug_img_with_noise \u001b[38;5;241m=\u001b[39m add_gaussian_noise(aug_img)\n",
      "File \u001b[1;32mc:\\Users\\User\\AppData\\Local\\Programs\\Python\\Python38\\lib\\site-packages\\keras\\preprocessing\\image.py:156\u001b[0m, in \u001b[0;36mIterator.__next__\u001b[1;34m(self, *args, **kwargs)\u001b[0m\n\u001b[0;32m    155\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21m__next__\u001b[39m(\u001b[38;5;28mself\u001b[39m, \u001b[38;5;241m*\u001b[39margs, \u001b[38;5;241m*\u001b[39m\u001b[38;5;241m*\u001b[39mkwargs):\n\u001b[1;32m--> 156\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mnext\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n",
      "File \u001b[1;32mc:\\Users\\User\\AppData\\Local\\Programs\\Python\\Python38\\lib\\site-packages\\keras\\preprocessing\\image.py:168\u001b[0m, in \u001b[0;36mIterator.next\u001b[1;34m(self)\u001b[0m\n\u001b[0;32m    165\u001b[0m     index_array \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mnext\u001b[39m(\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mindex_generator)\n\u001b[0;32m    166\u001b[0m \u001b[38;5;66;03m# The transformation of images is not under thread lock\u001b[39;00m\n\u001b[0;32m    167\u001b[0m \u001b[38;5;66;03m# so it can be done in parallel\u001b[39;00m\n\u001b[1;32m--> 168\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_get_batches_of_transformed_samples\u001b[49m\u001b[43m(\u001b[49m\u001b[43mindex_array\u001b[49m\u001b[43m)\u001b[49m\n",
      "File \u001b[1;32mc:\\Users\\User\\AppData\\Local\\Programs\\Python\\Python38\\lib\\site-packages\\keras\\preprocessing\\image.py:801\u001b[0m, in \u001b[0;36mNumpyArrayIterator._get_batches_of_transformed_samples\u001b[1;34m(self, index_array)\u001b[0m\n\u001b[0;32m    799\u001b[0m x \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mx[j]\n\u001b[0;32m    800\u001b[0m params \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mimage_data_generator\u001b[38;5;241m.\u001b[39mget_random_transform(x\u001b[38;5;241m.\u001b[39mshape)\n\u001b[1;32m--> 801\u001b[0m x \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mimage_data_generator\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mapply_transform\u001b[49m\u001b[43m(\u001b[49m\n\u001b[0;32m    802\u001b[0m \u001b[43m    \u001b[49m\u001b[43mx\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mastype\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mdtype\u001b[49m\u001b[43m)\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mparams\u001b[49m\n\u001b[0;32m    803\u001b[0m \u001b[43m\u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m    804\u001b[0m x \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mimage_data_generator\u001b[38;5;241m.\u001b[39mstandardize(x)\n\u001b[0;32m    805\u001b[0m batch_x[i] \u001b[38;5;241m=\u001b[39m x\n",
      "File \u001b[1;32mc:\\Users\\User\\AppData\\Local\\Programs\\Python\\Python38\\lib\\site-packages\\keras\\preprocessing\\image.py:2011\u001b[0m, in \u001b[0;36mImageDataGenerator.apply_transform\u001b[1;34m(self, x, transform_parameters)\u001b[0m\n\u001b[0;32m   2008\u001b[0m img_col_axis \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mcol_axis \u001b[38;5;241m-\u001b[39m \u001b[38;5;241m1\u001b[39m\n\u001b[0;32m   2009\u001b[0m img_channel_axis \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mchannel_axis \u001b[38;5;241m-\u001b[39m \u001b[38;5;241m1\u001b[39m\n\u001b[1;32m-> 2011\u001b[0m x \u001b[38;5;241m=\u001b[39m \u001b[43mapply_affine_transform\u001b[49m\u001b[43m(\u001b[49m\n\u001b[0;32m   2012\u001b[0m \u001b[43m    \u001b[49m\u001b[43mx\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m   2013\u001b[0m \u001b[43m    \u001b[49m\u001b[43mtransform_parameters\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mget\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43mtheta\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m0\u001b[39;49m\u001b[43m)\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m   2014\u001b[0m \u001b[43m    \u001b[49m\u001b[43mtransform_parameters\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mget\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43mtx\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m0\u001b[39;49m\u001b[43m)\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m   2015\u001b[0m \u001b[43m    \u001b[49m\u001b[43mtransform_parameters\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mget\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43mty\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m0\u001b[39;49m\u001b[43m)\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m   2016\u001b[0m \u001b[43m    \u001b[49m\u001b[43mtransform_parameters\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mget\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43mshear\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m0\u001b[39;49m\u001b[43m)\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m   2017\u001b[0m \u001b[43m    \u001b[49m\u001b[43mtransform_parameters\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mget\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43mzx\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m1\u001b[39;49m\u001b[43m)\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m   2018\u001b[0m \u001b[43m    \u001b[49m\u001b[43mtransform_parameters\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mget\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43mzy\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m1\u001b[39;49m\u001b[43m)\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m   2019\u001b[0m \u001b[43m    \u001b[49m\u001b[43mrow_axis\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mimg_row_axis\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m   2020\u001b[0m \u001b[43m    \u001b[49m\u001b[43mcol_axis\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mimg_col_axis\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m   2021\u001b[0m \u001b[43m    \u001b[49m\u001b[43mchannel_axis\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mimg_channel_axis\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m   2022\u001b[0m \u001b[43m    \u001b[49m\u001b[43mfill_mode\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mfill_mode\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m   2023\u001b[0m \u001b[43m    \u001b[49m\u001b[43mcval\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mcval\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m   2024\u001b[0m \u001b[43m    \u001b[49m\u001b[43morder\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43minterpolation_order\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m   2025\u001b[0m \u001b[43m\u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m   2027\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m transform_parameters\u001b[38;5;241m.\u001b[39mget(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mchannel_shift_intensity\u001b[39m\u001b[38;5;124m\"\u001b[39m) \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n\u001b[0;32m   2028\u001b[0m     x \u001b[38;5;241m=\u001b[39m apply_channel_shift(\n\u001b[0;32m   2029\u001b[0m         x,\n\u001b[0;32m   2030\u001b[0m         transform_parameters[\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mchannel_shift_intensity\u001b[39m\u001b[38;5;124m\"\u001b[39m],\n\u001b[0;32m   2031\u001b[0m         img_channel_axis,\n\u001b[0;32m   2032\u001b[0m     )\n",
      "File \u001b[1;32mc:\\Users\\User\\AppData\\Local\\Programs\\Python\\Python38\\lib\\site-packages\\keras\\preprocessing\\image.py:2604\u001b[0m, in \u001b[0;36mapply_affine_transform\u001b[1;34m(x, theta, tx, ty, shear, zx, zy, row_axis, col_axis, channel_axis, fill_mode, cval, order)\u001b[0m\n\u001b[0;32m   2594\u001b[0m \u001b[38;5;66;03m# Matrix construction assumes that coordinates are x, y (in that order).\u001b[39;00m\n\u001b[0;32m   2595\u001b[0m \u001b[38;5;66;03m# However, regular numpy arrays use y,x (aka i,j) indexing.\u001b[39;00m\n\u001b[0;32m   2596\u001b[0m \u001b[38;5;66;03m# Possible solution is:\u001b[39;00m\n\u001b[1;32m   (...)\u001b[0m\n\u001b[0;32m   2601\u001b[0m \u001b[38;5;66;03m# M' = PMP, where P is the permutation matrix, M is the original\u001b[39;00m\n\u001b[0;32m   2602\u001b[0m \u001b[38;5;66;03m# transformation matrix.\u001b[39;00m\n\u001b[0;32m   2603\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m col_axis \u001b[38;5;241m>\u001b[39m row_axis:\n\u001b[1;32m-> 2604\u001b[0m     \u001b[43mtransform_matrix\u001b[49m\u001b[43m[\u001b[49m\u001b[43m:\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43m[\u001b[49m\u001b[38;5;241;43m0\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m1\u001b[39;49m\u001b[43m]\u001b[49m\u001b[43m]\u001b[49m \u001b[38;5;241m=\u001b[39m transform_matrix[:, [\u001b[38;5;241m1\u001b[39m, \u001b[38;5;241m0\u001b[39m]]\n\u001b[0;32m   2605\u001b[0m     transform_matrix[[\u001b[38;5;241m0\u001b[39m, \u001b[38;5;241m1\u001b[39m]] \u001b[38;5;241m=\u001b[39m transform_matrix[[\u001b[38;5;241m1\u001b[39m, \u001b[38;5;241m0\u001b[39m]]\n\u001b[0;32m   2606\u001b[0m final_affine_matrix \u001b[38;5;241m=\u001b[39m transform_matrix[:\u001b[38;5;241m2\u001b[39m, :\u001b[38;5;241m2\u001b[39m]\n",
      "\u001b[1;31mKeyboardInterrupt\u001b[0m: "
     ]
    }
   ],
   "source": [
    "img = images[0]\n",
    "\n",
    "print(\"Min pixel value:\", np.min(img))\n",
    "print(\"Max pixel value:\", np.max(img))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data Preparation\n",
    "\n",
    "In this section, we will prepare the dataset for training our deep learning model. The steps involved are:\n",
    "\n",
    "1. **Extracting Letters from CAPTCHA Images**: Using the pre-trained YOLO model, we will extract individual letters from the CAPTCHA images.\n",
    "2. **Storing Extracted Letters and Labels**: We will store the extracted letters and their corresponding labels in separate lists.\n",
    "3. **Normalizing the Images**: The extracted letter images will be normalized to have pixel values between 0 and 1.\n",
    "\n",
    "### Steps:\n",
    "\n",
    "1. **Extract Letters**: Iterate through each CAPTCHA image file, use the `extract_letters_from_captcha` function to extract letters, and store them in the `images` list. The corresponding labels will be stored in the `labels` list.\n",
    "2. **Normalize Images**: Convert the `images` list to a NumPy array and normalize the pixel values to be between 0 and 1.\n",
    "\n",
    "By the end of this section, we will have a prepared dataset of letter images and their labels, ready for training our deep learning model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "112540\n",
      "112540\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
    "\n",
    "images_uint8 = np.array(images * 255, dtype=np.uint8)\n",
    "\n",
    "\n",
    "datagen = ImageDataGenerator(\n",
    "    rotation_range=10,     \n",
    "    width_shift_range=0.05,      \n",
    "    height_shift_range=0.05,     \n",
    "    zoom_range=0.1,             \n",
    "    shear_range=0.2,          \n",
    "    brightness_range=[0.5, 1.5], \n",
    "    fill_mode='constant',        \n",
    "    cval=170                 \n",
    ")\n",
    "\n",
    "def add_gaussian_noise(image, mean=0, std=8):\n",
    "    noise = np.random.normal(mean, std, image.shape)  \n",
    "    noisy_image = image + noise \n",
    "    noisy_image = np.clip(noisy_image, 0, 255) \n",
    "    return noisy_image.astype(np.uint8)\n",
    "\n",
    "\n",
    "augmented_images = []\n",
    "augmented_labels = []\n",
    "\n",
    "for i, img in enumerate(images_uint8):\n",
    "    img = np.expand_dims(img, axis=-1) \n",
    "    img = np.expand_dims(img, axis=0)\n",
    "\n",
    "    for _ in range(20):\n",
    "        aug_img = next(datagen.flow(img, batch_size=1))[0].astype(np.uint8)\n",
    "        aug_img = aug_img.squeeze()\n",
    "        aug_img_with_noise = add_gaussian_noise(aug_img)\n",
    "\n",
    "        augmented_images.append(aug_img_with_noise)\n",
    "        augmented_labels.append(labels[i])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "images shape :  (5627, 52, 32)\n",
      "labels shape :  (5627,)\n"
     ]
    }
   ],
   "source": [
    "print(len(augmented_images))\n",
    "print(len(augmented_labels))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'2': 286, '3': 273, '4': 299, '5': 306, '6': 293, '7': 285, '8': 291, 'a': 31, 'b': 256, 'c': 296, 'd': 278, 'e': 262, 'f': 281, 'g': 291, 'm': 265, 'n': 548, 'p': 285, 'w': 260, 'x': 287, 'y': 254}\n"
     ]
    }
   ],
   "source": [
    "print(\"images shape : \" , images.shape)\n",
    "print(\"labels shape : \" , labels.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(118167, 52, 32)\n",
      "(118167,)\n"
     ]
    }
   ],
   "source": [
    "\n",
    "x = np.concatenate((images, augmented_images), axis=0)\n",
    "y = np.concatenate((labels, augmented_labels), axis=0)\n",
    "\n",
    "print(x.shape) \n",
    "print(y.shape) \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "show_image(x[100857])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Label classes: ['2' '3' '4' '5' '6' '7' '8' 'a' 'b' 'c' 'd' 'e' 'f' 'g' 'm' 'n' 'p' 'w' 'x' 'y']\n",
      "Sample encoded labels: [ 0  0  4 14 10]\n",
      "Sample one-hot labels: [[          1           0           0           0           0           0           0           0           0           0           0           0           0           0           0           0           0           0           0           0]\n",
      " [          1           0           0           0           0           0           0           0           0           0           0           0           0           0           0           0           0           0           0           0]\n",
      " [          0           0           0           0           1           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           0           0           0           0           1           0           0           0           0           0]\n",
      " [          0           0           0           0           0           0           0           0           0           0           1           0           0           0           0           0           0           0           0           0]]\n"
     ]
    }
   ],
   "source": [
    "label_encoder = LabelEncoder()\n",
    "encoded_labels = label_encoder.fit_transform(y)\n",
    "\n",
    "one_hot_labels = to_categorical(encoded_labels)\n",
    "\n",
    "print(\"Label classes:\", label_encoder.classes_)\n",
    "print(\"Sample encoded labels:\", encoded_labels[:5])\n",
    "print(\"Sample one-hot labels:\", one_hot_labels[:5])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['6'], dtype='<U1')"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "label_encoder.inverse_transform([one_hot_labels[100857].argmax()])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.save('x.npy', x)\n",
    "np.save('one_hot_lables.npy', one_hot_labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_2\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " conv2d_6 (Conv2D)           (None, 52, 32, 32)        320       \n",
      "                                                                 \n",
      " max_pooling2d_6 (MaxPooling  (None, 26, 16, 32)       0         \n",
      " 2D)                                                             \n",
      "                                                                 \n",
      " batch_normalization_6 (Batc  (None, 26, 16, 32)       128       \n",
      " hNormalization)                                                 \n",
      "                                                                 \n",
      " dropout_8 (Dropout)         (None, 26, 16, 32)        0         \n",
      "                                                                 \n",
      " conv2d_7 (Conv2D)           (None, 26, 16, 64)        18496     \n",
      "                                                                 \n",
      " max_pooling2d_7 (MaxPooling  (None, 13, 8, 64)        0         \n",
      " 2D)                                                             \n",
      "                                                                 \n",
      " batch_normalization_7 (Batc  (None, 13, 8, 64)        256       \n",
      " hNormalization)                                                 \n",
      "                                                                 \n",
      " dropout_9 (Dropout)         (None, 13, 8, 64)         0         \n",
      "                                                                 \n",
      " conv2d_8 (Conv2D)           (None, 13, 8, 128)        73856     \n",
      "                                                                 \n",
      " max_pooling2d_8 (MaxPooling  (None, 6, 4, 128)        0         \n",
      " 2D)                                                             \n",
      "                                                                 \n",
      " batch_normalization_8 (Batc  (None, 6, 4, 128)        512       \n",
      " hNormalization)                                                 \n",
      "                                                                 \n",
      " dropout_10 (Dropout)        (None, 6, 4, 128)         0         \n",
      "                                                                 \n",
      " flatten_2 (Flatten)         (None, 3072)              0         \n",
      "                                                                 \n",
      " dense_4 (Dense)             (None, 128)               393344    \n",
      "                                                                 \n",
      " dropout_11 (Dropout)        (None, 128)               0         \n",
      "                                                                 \n",
      " dense_5 (Dense)             (None, 20)                2580      \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 489,492\n",
      "Trainable params: 489,044\n",
      "Non-trainable params: 448\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "import tensorflow as tf\n",
    "from tensorflow.keras.models import Sequential\n",
    "from tensorflow.keras.layers import Conv2D, MaxPooling2D, Dropout, Flatten, Dense, BatchNormalization\n",
    "\n",
    "def create_captcha_model(input_shape, num_classes):\n",
    "    model = Sequential([\n",
    "        Conv2D(32, (3, 3), activation='relu', input_shape=input_shape, padding='same'),\n",
    "        MaxPooling2D((2, 2)),\n",
    "        BatchNormalization(),\n",
    "        Dropout(0.5),\n",
    "        Conv2D(64, (3, 3), activation='relu', padding='same'),\n",
    "        MaxPooling2D((2, 2)),\n",
    "        BatchNormalization(),\n",
    "        Dropout(0.5),\n",
    "\n",
    "        Conv2D(128, (3, 3), activation='relu', padding='same'),\n",
    "        MaxPooling2D((2, 2)),\n",
    "        BatchNormalization(),\n",
    "        Dropout(0.5),\n",
    "\n",
    "        Flatten(),\n",
    "        Dense(128, activation='relu'),\n",
    "        Dropout(0.5),\n",
    "        Dense(num_classes, activation='softmax')\n",
    "    ])\n",
    "\n",
    "    model.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy'])\n",
    "    return model\n",
    "\n",
    "model = create_captcha_model((52, 32, 1), one_hot_labels.shape[1])\n",
    "\n",
    "model.summary()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/30\n",
      "6753/6753 [==============================] - 113s 17ms/step - loss: 0.6234 - accuracy: 0.8213 - val_loss: 0.2857 - val_accuracy: 0.9605\n",
      "Epoch 2/30\n",
      "6753/6753 [==============================] - 115s 17ms/step - loss: 0.2570 - accuracy: 0.9276 - val_loss: 0.2726 - val_accuracy: 0.9669\n",
      "Epoch 3/30\n",
      "6753/6753 [==============================] - 116s 17ms/step - loss: 0.1946 - accuracy: 0.9431 - val_loss: 0.2709 - val_accuracy: 0.9659\n",
      "Epoch 4/30\n",
      "6753/6753 [==============================] - 116s 17ms/step - loss: 0.1644 - accuracy: 0.9525 - val_loss: 0.2684 - val_accuracy: 0.9695\n",
      "Epoch 5/30\n",
      "6753/6753 [==============================] - 117s 17ms/step - loss: 0.1312 - accuracy: 0.9598 - val_loss: 0.2544 - val_accuracy: 0.9723\n",
      "Epoch 6/30\n",
      "6753/6753 [==============================] - 116s 17ms/step - loss: 0.1177 - accuracy: 0.9642 - val_loss: 0.2493 - val_accuracy: 0.9735\n",
      "Epoch 7/30\n",
      "6753/6753 [==============================] - 118s 17ms/step - loss: 0.1090 - accuracy: 0.9673 - val_loss: 0.2214 - val_accuracy: 0.9735\n",
      "Epoch 8/30\n",
      "6753/6753 [==============================] - 116s 17ms/step - loss: 0.0954 - accuracy: 0.9701 - val_loss: 0.2258 - val_accuracy: 0.9747\n",
      "Epoch 9/30\n",
      "6753/6753 [==============================] - 117s 17ms/step - loss: 0.0908 - accuracy: 0.9717 - val_loss: 0.2329 - val_accuracy: 0.9744\n",
      "Epoch 10/30\n",
      "6753/6753 [==============================] - 119s 18ms/step - loss: 0.0853 - accuracy: 0.9737 - val_loss: 0.2454 - val_accuracy: 0.9718\n",
      "Epoch 11/30\n",
      "6753/6753 [==============================] - 119s 18ms/step - loss: 0.0849 - accuracy: 0.9745 - val_loss: 0.2646 - val_accuracy: 0.9756\n",
      "Epoch 12/30\n",
      "6753/6753 [==============================] - 119s 18ms/step - loss: 0.0818 - accuracy: 0.9744 - val_loss: 0.2331 - val_accuracy: 0.9785\n"
     ]
    }
   ],
   "source": [
    "from sklearn.utils.class_weight import compute_class_weight\n",
    "import numpy as np\n",
    "\n",
    "class_weights = compute_class_weight(\n",
    "    class_weight='balanced',\n",
    "    classes=np.unique(labels), \n",
    "    y=labels\n",
    ")\n",
    "\n",
    "\n",
    "early_stopping = EarlyStopping(\n",
    "    monitor='val_loss', \n",
    "    patience=5,     \n",
    "    restore_best_weights=True \n",
    "\n",
    "\n",
    "history = model.fit(\n",
    "    x, one_hot_labels,\n",
    "    batch_size=14,\n",
    "    validation_split=0.2,\n",
    "    epochs=30,\n",
    "    class_weight=dict(enumerate(class_weights)),\n",
    "    callbacks=[early_stopping] \n",
    ")\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(history.history['accuracy'], label='Accuracy')\n",
    "plt.plot(history.history['val_accuracy'], label='Validation Accuracy')\n",
    "plt.xlabel('Epoch')\n",
    "plt.ylabel('Accuracy')\n",
    "plt.legend(loc='lower right')\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "show_image(images[477])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "labels[477]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model saved to BESTREAL.h5\n"
     ]
    }
   ],
   "source": [
    "#model.save('FINALMODELS/mymodel.h5')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "# Assuming `label_encoder` is already trained\n",
    "#joblib.dump(label_encoder, 'trained_label_encoder.pkl')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "image number : 412\n",
      "\n",
      "image 1/1 c:\\Users\\User\\Desktop\\cap\\captchas\\captcha_412.png: 128x640 8 characterss, 58.8ms\n",
      "Speed: 7.8ms preprocess, 58.8ms inference, 0.2ms postprocess per image at shape (1, 3, 128, 640)\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": [
      "Number of letters extracted: 8\n",
      "1/1 [==============================] - 0s 156ms/step\n",
      "1/1 [==============================] - 0s 14ms/step\n",
      "1/1 [==============================] - 0s 13ms/step\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "1/1 [==============================] - 0s 14ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 17ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "n4pnbepn\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x1000 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "rendnum = random.choice(range(30 , 500))\n",
    "image_path_test = f\"captchas/captcha_{rendnum}.png\"\n",
    "print(f\"image number : {rendnum}\")\n",
    "image_test = cv2.imread(image_path_test)\n",
    "model_test = YOLO(\"yolo_model2/best.pt\")\n",
    "results_test = model_test(image_path_test)\n",
    "detections_test = results_test[0].boxes\n",
    "detections_test = sorted(detections_test, key=lambda box: box.xyxy[0][0])\n",
    "\n",
    "def check_letter_tarakom(pic):\n",
    "    _, pic = cv2.threshold(pic, 127, 255, cv2.THRESH_BINARY)\n",
    "    s = 90 - (np.sum(pic , axis=0 , keepdims=True) / 255)\n",
    "    total = len(s[0])\n",
    "    howmanyblack = 0\n",
    "    for i in s[0]:\n",
    "        if np.sum(i) >= 175:\n",
    "            howmanyblack += 1\n",
    "    if total - howmanyblack <= 22:\n",
    "        return True\n",
    "    else:\n",
    "        return False\n",
    "\n",
    "letters_test = []\n",
    "\n",
    "last_one = []\n",
    "\n",
    "howmany = 0 \n",
    "for box in detections_test:\n",
    "    x1, y1, x2, y2 = map(int, box.xyxy[0].tolist())\n",
    "    if last_one != []:\n",
    "        if int(x1) - last_one[0] > 10:\n",
    "            letter_crop = image_test[y1:y2, x1:x2] \n",
    "            resized_letter = cv2.resize(letter_crop, (32, 52)) \n",
    "            resized_letter = cv2.cvtColor(resized_letter, cv2.COLOR_BGR2GRAY) \n",
    "            if howmany == 8:\n",
    "                if check_letter_tarakom(letter_crop) == True:\n",
    "                        _, binarized = cv2.threshold(resized_letter, 128, 255, cv2.THRESH_BINARY)\n",
    "                        letters_test.append(resized_letter)\n",
    "            elif howmany <= 7:\n",
    "                    _, binarized = cv2.threshold(resized_letter, 128, 255, cv2.THRESH_BINARY)\n",
    "                    letters_test.append(resized_letter)\n",
    "            last_one = [int(x1), int(y1), int(x2), int(y2)]\n",
    "            howmany += 1\n",
    "    else:\n",
    "        letter_crop = image_test[y1:y2, x1:x2] \n",
    "        resized_letter = cv2.resize(letter_crop, (32, 52)) \n",
    "        resized_letter = cv2.cvtColor(resized_letter, cv2.COLOR_BGR2GRAY)\n",
    "        if check_letter_tarakom(letter_crop) == True:\n",
    "                _, binarized = cv2.threshold(resized_letter, 128, 255, cv2.THRESH_BINARY)\n",
    "                letters_test.append(resized_letter)\n",
    "        last_one = [int(x1), int(y1), int(x2), int(y2)]\n",
    "        howmany += 1\n",
    "\n",
    "show_image(image_test)\n",
    "print(f\"Number of letters extracted: {len(letters_test)}\")\n",
    "\n",
    "allpredicted = []\n",
    "\n",
    "for i in range(0,len(letters_test)):\n",
    "    sample_image = np.array(letters_test[i])  \n",
    "    sample_image = sample_image.reshape(1, 52, 32, 1) \n",
    "    predicted_label = model.predict(sample_image)\n",
    "    predicted_class = label_encoder.inverse_transform([predicted_label.argmax()])\n",
    "    allpredicted.append(predicted_class[0])\n",
    "captcha_result = \"\".join(allpredicted)\n",
    "print(captcha_result)\n",
    "#-------------------\n",
    "last_one = []\n",
    "\n",
    "def add_to_dataset(add:False , allpredicted , lab:str):\n",
    "    global image_test\n",
    "    tosave = image_test.copy()\n",
    "    if add == True:\n",
    "        if f\"kaggledataset/{lab}.png\" not in os.listdir(\"kaggledataset\"):\n",
    "            cv2.imwrite(f\"kaggledataset/{lab}.png\" , tosave)\n",
    "            print(\"saved\")\n",
    "        else:\n",
    "            print(\"existed\")\n",
    "for_show = image_test.copy()\n",
    "\n",
    "howmany = 0 \n",
    "for box in detections_test:\n",
    "    try:\n",
    "        x1, y1, x2, y2 = box.xyxy[0].tolist() \n",
    "        if last_one != []:\n",
    "            if  int(x1) - last_one[0] >= 11:\n",
    "                confidence = box.conf[0] \n",
    "                cls = int(box.cls[0]) \n",
    "                x1, y1, x2, y2 = map(int, box.xyxy[0].tolist())\n",
    "                ff = for_show[y1:y2, x1:x2]\n",
    "                if howmany == 8:\n",
    "                    if check_letter_tarakom(ff) == True:\n",
    "                        cv2.rectangle(for_show, (int(x1), int(y1)), (int(x2), int(y2)), color=(255, 0, 0), thickness=1)\n",
    "                elif howmany <= 7:\n",
    "                    cv2.rectangle(for_show, (int(x1), int(y1)), (int(x2), int(y2)), color=(255, 0, 0), thickness=1)\n",
    "                \n",
    "                labelr = f\"Class {cls} ({confidence:.2f})\"\n",
    "                cv2.putText(for_show, labelr, (int(x1), int(y1) - 10), cv2.FONT_HERSHEY_SIMPLEX, 0.5, (255, 0, 0), 2)\n",
    "                last_one = [int(x1), int(y1), int(x2), int(y2)]\n",
    "                howmany +=1\n",
    "        else:\n",
    "            x1, y1, x2, y2 = map(int, box.xyxy[0].tolist())\n",
    "            ff = for_show[y1:y2, x1:x2]\n",
    "            if check_letter_tarakom(ff) == True:\n",
    "                cv2.rectangle(for_show, (int(x1), int(y1)), (int(x2), int(y2)), color=(255, 0, 0), thickness=1)\n",
    "            last_one = [int(x1), int(y1), int(x2), int(y2)]\n",
    "            howmany +=1\n",
    "    except Exception as e:\n",
    "        print(e)\n",
    "        x1, y1, x2, y2 = map(int, box.xyxy[0].tolist())\n",
    "        ff = for_show[y1:y2, x1:x2]\n",
    "        cv2.rectangle(for_show, (int(x1), int(y1)), (int(x2), int(y2)), color=(255, 0, 0), thickness=1)\n",
    "        last_one = [int(x1), int(y1), int(x2), int(y2)]\n",
    "        howmany += 1\n",
    "\n",
    "plt.figure(figsize=(10, 10))  \n",
    "plt.imshow(for_show)\n",
    "plt.show()\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "images :  (118167, 52, 32)\n",
      "labels :  (118167,)\n"
     ]
    }
   ],
   "source": [
    "\n",
    "rendnum = random.choice(range(30 , 500))\n",
    "image_path_test = f\"captchas/captcha_{rendnum}.png\"\n",
    "print(f\"image number : {rendnum}\")\n",
    "image_test = cv2.imread(image_path_test)\n",
    "model_test = YOLO(\"yolo_model2/best.pt\")\n",
    "results_test = model_test(image_path_test)\n",
    "detections_test = results_test[0].boxes\n",
    "detections_test = sorted(detections_test, key=lambda box: box.xyxy[0][0])\n",
    "\n",
    "def check_letter_tarakom(pic):\n",
    "    _, pic = cv2.threshold(pic, 127, 255, cv2.THRESH_BINARY)\n",
    "    s = 90 - (np.sum(pic , axis=0 , keepdims=True) / 255)\n",
    "    total = len(s[0])\n",
    "    howmanyblack = 0\n",
    "    for i in s[0]:\n",
    "        if np.sum(i) >= 175:\n",
    "            howmanyblack += 1\n",
    "    if total - howmanyblack <= 22:\n",
    "        return True\n",
    "    else:\n",
    "        return False\n",
    "\n",
    "letters_test = []\n",
    "\n",
    "last_one = []\n",
    "\n",
    "howmany = 0 \n",
    "for box in detections_test:\n",
    "    x1, y1, x2, y2 = map(int, box.xyxy[0].tolist())\n",
    "    if last_one != []:\n",
    "        if int(x1) - last_one[0] > 10:\n",
    "            letter_crop = image_test[y1:y2, x1:x2] \n",
    "            resized_letter = cv2.resize(letter_crop, (32, 52)) \n",
    "            resized_letter = cv2.cvtColor(resized_letter, cv2.COLOR_BGR2GRAY) \n",
    "            if howmany == 8:\n",
    "                if check_letter_tarakom(letter_crop) == True:\n",
    "                        _, binarized = cv2.threshold(resized_letter, 128, 255, cv2.THRESH_BINARY)\n",
    "                        letters_test.append(resized_letter)\n",
    "            elif howmany <= 7:\n",
    "                    _, binarized = cv2.threshold(resized_letter, 128, 255, cv2.THRESH_BINARY)\n",
    "                    letters_test.append(resized_letter)\n",
    "            last_one = [int(x1), int(y1), int(x2), int(y2)]\n",
    "            howmany += 1\n",
    "    else:\n",
    "        letter_crop = image_test[y1:y2, x1:x2] \n",
    "        resized_letter = cv2.resize(letter_crop, (32, 52)) \n",
    "        resized_letter = cv2.cvtColor(resized_letter, cv2.COLOR_BGR2GRAY)\n",
    "        if check_letter_tarakom(letter_crop) == True:\n",
    "                _, binarized = cv2.threshold(resized_letter, 128, 255, cv2.THRESH_BINARY)\n",
    "                letters_test.append(resized_letter)\n",
    "        last_one = [int(x1), int(y1), int(x2), int(y2)]\n",
    "        howmany += 1\n",
    "\n",
    "show_image(image_test)\n",
    "print(f\"Number of letters extracted: {len(letters_test)}\")\n",
    "\n",
    "allpredicted = []\n",
    "\n",
    "for i in range(0,len(letters_test)):\n",
    "    sample_image = np.array(letters_test[i])  \n",
    "    sample_image = sample_image.reshape(1, 52, 32, 1) \n",
    "    predicted_label = model.predict(sample_image)\n",
    "    predicted_class = label_encoder.inverse_transform([predicted_label.argmax()])\n",
    "    allpredicted.append(predicted_class[0])\n",
    "captcha_result = \"\".join(allpredicted)\n",
    "print(captcha_result)\n",
    "#-------------------\n",
    "last_one = []\n",
    "\n",
    "def add_to_dataset(add:False , allpredicted , lab:str):\n",
    "    global image_test\n",
    "    tosave = image_test.copy()\n",
    "    if add == True:\n",
    "        if f\"kaggledataset/{lab}.png\" not in os.listdir(\"kaggledataset\"):\n",
    "            cv2.imwrite(f\"kaggledataset/{lab}.png\" , tosave)\n",
    "            print(\"saved\")\n",
    "        else:\n",
    "            print(\"existed\")\n",
    "for_show = image_test.copy()\n",
    "\n",
    "howmany = 0 \n",
    "for box in detections_test:\n",
    "    try:\n",
    "        x1, y1, x2, y2 = box.xyxy[0].tolist() \n",
    "        if last_one != []:\n",
    "            if  int(x1) - last_one[0] >= 11:\n",
    "                confidence = box.conf[0] \n",
    "                cls = int(box.cls[0]) \n",
    "                x1, y1, x2, y2 = map(int, box.xyxy[0].tolist())\n",
    "                ff = for_show[y1:y2, x1:x2]\n",
    "                if howmany == 8:\n",
    "                    if check_letter_tarakom(ff) == True:\n",
    "                        cv2.rectangle(for_show, (int(x1), int(y1)), (int(x2), int(y2)), color=(255, 0, 0), thickness=1)\n",
    "                elif howmany <= 7:\n",
    "                    cv2.rectangle(for_show, (int(x1), int(y1)), (int(x2), int(y2)), color=(255, 0, 0), thickness=1)\n",
    "                \n",
    "                labelr = f\"Class {cls} ({confidence:.2f})\"\n",
    "                cv2.putText(for_show, labelr, (int(x1), int(y1) - 10), cv2.FONT_HERSHEY_SIMPLEX, 0.5, (255, 0, 0), 2)\n",
    "                last_one = [int(x1), int(y1), int(x2), int(y2)]\n",
    "                howmany +=1\n",
    "        else:\n",
    "            x1, y1, x2, y2 = map(int, box.xyxy[0].tolist())\n",
    "            ff = for_show[y1:y2, x1:x2]\n",
    "            if check_letter_tarakom(ff) == True:\n",
    "                cv2.rectangle(for_show, (int(x1), int(y1)), (int(x2), int(y2)), color=(255, 0, 0), thickness=1)\n",
    "            last_one = [int(x1), int(y1), int(x2), int(y2)]\n",
    "            howmany +=1\n",
    "    except Exception as e:\n",
    "        print(e)\n",
    "        x1, y1, x2, y2 = map(int, box.xyxy[0].tolist())\n",
    "        ff = for_show[y1:y2, x1:x2]\n",
    "        cv2.rectangle(for_show, (int(x1), int(y1)), (int(x2), int(y2)), color=(255, 0, 0), thickness=1)\n",
    "        last_one = [int(x1), int(y1), int(x2), int(y2)]\n",
    "        howmany += 1\n",
    "\n",
    "plt.figure(figsize=(10, 10))  \n",
    "plt.imshow(for_show)\n",
    "plt.show()\n",
    "\n",
    "\n"
   ]
  }
 ],
 "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": "3.8.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}