Driver Drowsiness Detection system _mini_model.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3f52c5f4",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "pygame 2.0.1 (SDL 2.0.14, Python 3.8.5)\n",
      "Hello from the pygame community. https://www.pygame.org/contribute.html\n",
      "WARNING:tensorflow:From <ipython-input-1-724325017444>:53: Sequential.predict_classes (from tensorflow.python.keras.engine.sequential) is deprecated and will be removed after 2021-01-01.\n",
      "Instructions for updating:\n",
      "Please use instead:* `np.argmax(model.predict(x), axis=-1)`,   if your model does multi-class classification   (e.g. if it uses a `softmax` last-layer activation).* `(model.predict(x) > 0.5).astype(\"int32\")`,   if your model does binary classification   (e.g. if it uses a `sigmoid` last-layer activation).\n"
     ]
    }
   ],
   "source": [
    "import cv2\n",
    "import os\n",
    "from keras.models import load_model\n",
    "import numpy as np\n",
    "from pygame import mixer\n",
    "import time\n",
    "\n",
    "\n",
    "mixer.init()\n",
    "sound = mixer.Sound('alarm.wav')\n",
    "\n",
    "face = cv2.CascadeClassifier('haar cascade files\\haarcascade_frontalface_alt.xml')\n",
    "leye = cv2.CascadeClassifier('haar cascade files\\haarcascade_lefteye_2splits.xml')\n",
    "reye = cv2.CascadeClassifier('haar cascade files\\haarcascade_righteye_2splits.xml')\n",
    "\n",
    "\n",
    "\n",
    "lbl=['Close','Open']\n",
    "\n",
    "model = load_model('models/cnncat2.h5')\n",
    "path = os.getcwd()\n",
    "cap = cv2.VideoCapture(0)\n",
    "font = cv2.FONT_HERSHEY_COMPLEX_SMALL\n",
    "count=0\n",
    "score=0\n",
    "thicc=2\n",
    "rpred=[99]\n",
    "lpred=[99]\n",
    "\n",
    "while(True):\n",
    "    ret, frame = cap.read()\n",
    "    height,width = frame.shape[:2] \n",
    "\n",
    "    gray =cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)\n",
    "    \n",
    "    faces = face.detectMultiScale(gray,minNeighbors=5,scaleFactor=1.1,minSize=(25,25))\n",
    "    left_eye = leye.detectMultiScale(gray)\n",
    "    right_eye =  reye.detectMultiScale(gray)\n",
    "\n",
    "    cv2.rectangle(frame, (0,height-50) , (200,height) , (0,0,0) , thickness=cv2.FILLED )\n",
    "\n",
    "    for (x,y,w,h) in faces:\n",
    "        cv2.rectangle(frame, (x,y) , (x+w,y+h) , (100,100,100) , 1 )\n",
    "\n",
    "    for (x,y,w,h) in right_eye:\n",
    "        r_eye=frame[y:y+h,x:x+w]\n",
    "        count=count+1\n",
    "        r_eye = cv2.cvtColor(r_eye,cv2.COLOR_BGR2GRAY)\n",
    "        r_eye = cv2.resize(r_eye,(24,24))\n",
    "        r_eye= r_eye/255\n",
    "        r_eye=  r_eye.reshape(24,24,-1)\n",
    "        r_eye = np.expand_dims(r_eye,axis=0)\n",
    "        rpred = model.predict_classes(r_eye)\n",
    "        if(rpred[0]==1):\n",
    "            lbl='Open' \n",
    "        if(rpred[0]==0):\n",
    "            lbl='Closed'\n",
    "        break\n",
    "\n",
    "    for (x,y,w,h) in left_eye:\n",
    "        l_eye=frame[y:y+h,x:x+w]\n",
    "        count=count+1\n",
    "        l_eye = cv2.cvtColor(l_eye,cv2.COLOR_BGR2GRAY)  \n",
    "        l_eye = cv2.resize(l_eye,(24,24))\n",
    "        l_eye= l_eye/255\n",
    "        l_eye=l_eye.reshape(24,24,-1)\n",
    "        l_eye = np.expand_dims(l_eye,axis=0)\n",
    "        lpred = model.predict_classes(l_eye)\n",
    "        if(lpred[0]==1):\n",
    "            lbl='Open'   \n",
    "        if(lpred[0]==0):\n",
    "            lbl='Closed'\n",
    "        break\n",
    "\n",
    "    if(rpred[0]==0 and lpred[0]==0):\n",
    "        score=score+1\n",
    "        cv2.putText(frame,\"Closed\",(10,height-20), font, 1,(255,255,255),1,cv2.LINE_AA)\n",
    "    # if(rpred[0]==1 or lpred[0]==1):\n",
    "    else:\n",
    "        score=score-1\n",
    "        cv2.putText(frame,\"Open\",(10,height-20), font, 1,(255,255,255),1,cv2.LINE_AA)\n",
    "    \n",
    "        \n",
    "    if(score<0):\n",
    "        score=0   \n",
    "    cv2.putText(frame,'Score:'+str(score),(100,height-20), font, 1,(255,255,255),1,cv2.LINE_AA)\n",
    "    if(score>15):\n",
    "        #person is feeling sleepy so we beep the alarm\n",
    "        cv2.imwrite(os.path.join(path,'image.jpg'),frame)\n",
    "        try:\n",
    "            sound.play()\n",
    "            \n",
    "        except:  # isplaying = False\n",
    "            pass\n",
    "        if(thicc<16):\n",
    "            thicc= thicc+2\n",
    "        else:\n",
    "            thicc=thicc-2\n",
    "            if(thicc<2):\n",
    "                thicc=2\n",
    "        cv2.rectangle(frame,(0,0),(width,height),(0,0,255),thicc) \n",
    "    cv2.imshow('frame',frame)\n",
    "    if cv2.waitKey(1) & 0xFF == ord('q'):\n",
    "        break\n",
    "cap.release()\n",
    "cv2.destroyAllWindows()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e29e91c1",
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "from keras.preprocessing import image\n",
    "import matplotlib.pyplot as plt \n",
    "import numpy as np\n",
    "from keras.utils.np_utils import to_categorical\n",
    "import random,shutil\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dropout,Conv2D,Flatten,Dense, MaxPooling2D, BatchNormalization\n",
    "from keras.models import load_model\n",
    "\n",
    "\n",
    "def generator(dir, gen=image.ImageDataGenerator(rescale=1./255), shuffle=True,batch_size=1,target_size=(24,24),class_mode='categorical' ):\n",
    "\n",
    "    return gen.flow_from_directory(dir,batch_size=batch_size,shuffle=shuffle,color_mode='grayscale',class_mode=class_mode,target_size=target_size)\n",
    "\n",
    "BS= 32\n",
    "TS=(24,24)\n",
    "train_batch= generator('data/train',shuffle=True, batch_size=BS,target_size=TS)\n",
    "valid_batch= generator('data/valid',shuffle=True, batch_size=BS,target_size=TS)\n",
    "SPE= len(train_batch.classes)//BS\n",
    "VS = len(valid_batch.classes)//BS\n",
    "print(SPE,VS)\n",
    "\n",
    "\n",
    "# img,labels= next(train_batch)\n",
    "# print(img.shape)\n",
    "\n",
    "model = Sequential([\n",
    "    Conv2D(32, kernel_size=(3, 3), activation='relu', input_shape=(24,24,1)),\n",
    "    MaxPooling2D(pool_size=(1,1)),\n",
    "    Conv2D(32,(3,3),activation='relu'),\n",
    "    MaxPooling2D(pool_size=(1,1)),\n",
    "#32 convolution filters used each of size 3x3\n",
    "#again\n",
    "    Conv2D(64, (3, 3), activation='relu'),\n",
    "    MaxPooling2D(pool_size=(1,1)),\n",
    "\n",
    "#64 convolution filters used each of size 3x3\n",
    "#choose the best features via pooling\n",
    "    \n",
    "#randomly turn neurons on and off to improve convergence\n",
    "    Dropout(0.25),\n",
    "#flatten since too many dimensions, we only want a classification output\n",
    "    Flatten(),\n",
    "#fully connected to get all relevant data\n",
    "    Dense(128, activation='relu'),\n",
    "#one more dropout for convergence' sake :) \n",
    "    Dropout(0.5),\n",
    "#output a softmax to squash the matrix into output probabilities\n",
    "    Dense(2, activation='softmax')\n",
    "])\n",
    "\n",
    "model.compile(optimizer='adam',loss='categorical_crossentropy',metrics=['accuracy'])\n",
    "\n",
    "model.fit_generator(train_batch, validation_data=valid_batch,epochs=15,steps_per_epoch=SPE ,validation_steps=VS)\n",
    "\n",
    "model.save('models/cnnCat2.h5', overwrite=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "cc40cfee",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}