08_tracking_with_trackbar.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import cv2\n",
    "\n",
    "ESC_KEY = 27\n",
    "\n",
    "cap = cv2.VideoCapture(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "def setLimitsOfTrackbar():\n",
    "    hue = {}\n",
    "    hue[\"min\"] = cv2.getTrackbarPos(\"Min Hue\", trackbarWindow)\n",
    "    hue[\"max\"] = cv2.getTrackbarPos(\"Max Hue\", trackbarWindow)\n",
    "    \n",
    "    if hue[\"min\"] > hue[\"max\"]:\n",
    "        cv2.setTrackbarPos(\"Max Hue\", trackbarWindow, hue[\"min\"])\n",
    "        hue[\"max\"] = cv2.getTrackbarPos(\"Max Hue\", trackbarWindow)\n",
    "    \n",
    "    sat = {}\n",
    "    sat[\"min\"] = cv2.getTrackbarPos(\"Min Saturation\", trackbarWindow)\n",
    "    sat[\"max\"] = cv2.getTrackbarPos(\"Max Saturation\", trackbarWindow)\n",
    "    \n",
    "    if sat[\"min\"] > sat[\"max\"]:\n",
    "        cv2.setTrackbarPos(\"Max Saturation\", trackbarWindow, sat[\"min\"])\n",
    "        sat[\"max\"] = cv2.getTrackbarPos(\"Max Saturation\", trackbarWindow)\n",
    "\n",
    "    val = {}\n",
    "    val[\"min\"] = cv2.getTrackbarPos(\"Min Value\", trackbarWindow)\n",
    "    val[\"max\"] = cv2.getTrackbarPos(\"Max Value\", trackbarWindow)\n",
    "    \n",
    "    if val[\"min\"] > val[\"max\"]:\n",
    "        cv2.setTrackbarPos(\"Max Value\", trackbarWindow, val[\"min\"])\n",
    "        val[\"max\"] = cv2.getTrackbarPos(\"Max Value\", trackbarWindow)\n",
    "        \n",
    "    return hue, sat, val"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "def computeTracking(frame, hue, sat, val):\n",
    "    \n",
    "    #transforma a imagem de RGB para HSV\n",
    "    hsvImage = cv2.cvtColor(frame, cv2.COLOR_BGR2HSV)\n",
    "    \n",
    "    #definir os intervalos de cores que vão aparecer na imagem final\n",
    "    lowerColor = np.array([hue['min'], sat[\"min\"], val[\"min\"]])\n",
    "    upperColor = np.array([hue['max'], sat[\"max\"], val[\"max\"]])\n",
    "    \n",
    "    #marcador pra saber se o pixel pertence ao intervalo ou não\n",
    "    mask = cv2.inRange(hsvImage, lowerColor, upperColor)\n",
    "    \n",
    "    #aplica máscara que \"deixa passar\" pixels pertencentes ao intervalo, como filtro\n",
    "    result = cv2.bitwise_and(frame, frame, mask = mask)\n",
    "    \n",
    "    #aplica limiarização\n",
    "    gray = cv2.cvtColor(result, cv2.COLOR_BGR2GRAY)\n",
    "    _,gray = cv2.threshold(gray, 0, 255, cv2.THRESH_BINARY | cv2.THRESH_OTSU)\n",
    "    \n",
    "    #encontra pontos que circundam regiões conexas (contour)\n",
    "    contours, hierarchy = cv2.findContours(gray, cv2.RETR_LIST, cv2.CHAIN_APPROX_SIMPLE)\n",
    "    \n",
    "    #se existir contornos    \n",
    "    if contours:\n",
    "        #retornando a área do primeiro grupo de pixels brancos\n",
    "        maxArea = cv2.contourArea(contours[0])\n",
    "        contourMaxAreaId = 0\n",
    "        i = 0\n",
    "        \n",
    "        #para cada grupo de pixels branco\n",
    "        for cnt in contours:\n",
    "            #procura o grupo com a maior área\n",
    "            if maxArea < cv2.contourArea(cnt):\n",
    "                maxArea = cv2.contourArea(cnt)\n",
    "                contourMaxAreaId = i\n",
    "            i += 1\n",
    "            \n",
    "        #achei o contorno com maior área em pixels\n",
    "        cntMaxArea = contours[contourMaxAreaId]\n",
    "        \n",
    "        #retorna um retângulo que envolve o contorno em questão\n",
    "        xRect, yRect, wRect, hRect = cv2.boundingRect(cntMaxArea)\n",
    "        \n",
    "        #desenha caixa envolvente com espessura 3\n",
    "        cv2.rectangle(frame, (xRect, yRect), (xRect + wRect, yRect + hRect), (0, 0, 255), 2)\n",
    "    \n",
    "    return frame, gray"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "trackbarWindow = \"trackbar window\"\n",
    "cv2.namedWindow(trackbarWindow)\n",
    "\n",
    "def onChange(val):\n",
    "    return\n",
    "\n",
    "cv2.createTrackbar(\"Min Hue\", trackbarWindow, 0, 255, onChange)\n",
    "cv2.createTrackbar(\"Max Hue\", trackbarWindow, 255, 255, onChange)\n",
    "\n",
    "cv2.createTrackbar(\"Min Saturation\", trackbarWindow, 0, 255, onChange)\n",
    "cv2.createTrackbar(\"Max Saturation\", trackbarWindow, 255, 255, onChange)\n",
    "\n",
    "cv2.createTrackbar(\"Min Value\", trackbarWindow, 0, 255, onChange)\n",
    "cv2.createTrackbar(\"Max Value\", trackbarWindow, 255, 255, onChange)\n",
    "\n",
    "min_hue = cv2.getTrackbarPos(\"Min Hue\", trackbarWindow)\n",
    "max_hue = cv2.getTrackbarPos(\"Max Hue\", trackbarWindow)\n",
    "\n",
    "min_sat = cv2.getTrackbarPos(\"Min Saturation\", trackbarWindow)\n",
    "max_sat = cv2.getTrackbarPos(\"Max Saturation\", trackbarWindow)\n",
    "\n",
    "min_val = cv2.getTrackbarPos(\"Min Value\", trackbarWindow)\n",
    "max_val = cv2.getTrackbarPos(\"Max Value\", trackbarWindow)\n",
    "\n",
    "while True:\n",
    "    success, frame = cap.read()\n",
    "    \n",
    "    hue, sat, val = setLimitsOfTrackbar()\n",
    "    frame, gray = computeTracking(frame, hue, sat, val)\n",
    "    \n",
    "    cv2.imshow(\"mascara\", gray)\n",
    "    cv2.imshow(\"webcam\", frame)\n",
    "\n",
    "    if cv2.waitKey(1) & 0xFF == ord('q') or 0xFF == ESC_KEY:\n",
    "        break\n",
    "        \n",
    "cap.release()\n",
    "cv2.destroyAllWindows()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}