image_collector.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "8ccdd69e",
   "metadata": {},
   "outputs": [],
   "source": [
    "import cv2\n",
    "import uuid\n",
    "import mediapipe as mp"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "82bf9ecf",
   "metadata": {},
   "outputs": [],
   "source": [
    "camera = cv2.VideoCapture(0)\n",
    "def capture_image():\n",
    "    return_value,frame = camera.read()\n",
    "\n",
    "    return frame"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "17ae04b1",
   "metadata": {},
   "outputs": [],
   "source": [
    "mphands = mp.solutions.hands\n",
    "hands = mphands.Hands(max_num_hands=1)\n",
    "mp_drawing = mp.solutions.drawing_utils\n",
    "cap = cv2.VideoCapture(0)\n",
    "\n",
    "cap.set(cv2.CAP_PROP_FRAME_WIDTH, 1280)\n",
    "cap.set(cv2.CAP_PROP_FRAME_HEIGHT, 720)\n",
    "\n",
    "\n",
    "_, frame = cap.read()\n",
    "    \n",
    "\n",
    "h, w, c = frame.shape\n",
    "\n",
    "while True:\n",
    "    _, frame = cap.read()\n",
    "    frame= cv2.flip(frame,1)\n",
    "    framergb = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)\n",
    "    result = hands.process(framergb)\n",
    "    hand_landmarks = result.multi_hand_landmarks\n",
    "    if hand_landmarks:\n",
    "        for handLMs in hand_landmarks:\n",
    "            x_max = 0\n",
    "            y_max = 0\n",
    "            x_min = w\n",
    "            y_min = h\n",
    "            for lm in handLMs.landmark:\n",
    "                x, y = int(lm.x * w), int(lm.y * h)\n",
    "                if x > x_max:\n",
    "                    x_max = x\n",
    "                if x < x_min:\n",
    "                    x_min = x\n",
    "                if y > y_max:\n",
    "                    y_max = y\n",
    "                if y < y_min:\n",
    "                    y_min = y\n",
    "            cv2.rectangle(frame, (x_min-15, y_min-15), (x_max+15, y_max+15), (0, 255, 0), 2) \n",
    "            img_data=frame[y_min-15:y_max+15,x_min-15:x_max+15]\n",
    "            img_name ='Demo/Hands{}.jpg'.format(str(uuid.uuid1()))\n",
    "            cv2.imwrite(img_name, img_data)\n",
    "    cv2.imshow(\"Hand Detection\", frame)\n",
    "    if cv2.waitKey(1) & 0xFF==ord('q'):\n",
    "        break\n",
    "cv2.destroyAllWindows()        \n",
    "cap.release()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "74ce7c75",
   "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": "3.8.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}