Untitled.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import cv2\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "125 292\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "orig = cv2.imread(r'C:\\Users\\Legion\\Documents\\n2.jpg',0)\n",
    "scale = 8\n",
    "dim_x,dim_y = orig.shape\n",
    "print(dim_x//scale,dim_y//scale)\n",
    "resized_im = cv2.resize(orig,( dim_y//scale, dim_x//scale)) \n",
    "\n",
    "#calc hist\n",
    "histogram = cv2.calcHist([resized_im],[0],None,[256],[0,255])\n",
    "\n",
    "plt.plot(histogram)\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "-1"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cv2.imshow('resized',resized_im)\n",
    "cv2.waitKey(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "127.0\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "-1"
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#threshold\n",
    "retval,thres=cv2.threshold(resized_im,127,255,cv2.THRESH_BINARY_INV)\n",
    "\n",
    "print(retval)\n",
    "\n",
    "cv2.imshow('thresholded',thres)\n",
    "cv2.waitKey(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [],
   "source": [
    "kernel_length = np.array(orig).shape[1]//80\n",
    "\n",
    "vertical_kernel = cv2.getStructuringElement(cv2.MORPH_RECT, (1,kernel_length))\n",
    "\n",
    "horizontal_kernel = cv2.getStructuringElement(cv2.MORPH_RECT, (kernel_length,1))\n",
    "\n",
    "kernel = cv2.getStructuringElement(cv2.MORPH_RECT, (3,3))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Help on built-in function erode:\n",
      "\n",
      "erode(...)\n",
      "    erode(src, kernel[, dst[, anchor[, iterations[, borderType[, borderValue]]]]]) -> dst\n",
      "    .   @brief Erodes an image by using a specific structuring element.\n",
      "    .   \n",
      "    .   The function erodes the source image using the specified structuring element that determines the\n",
      "    .   shape of a pixel neighborhood over which the minimum is taken:\n",
      "    .   \n",
      "    .   \\f[\\texttt{dst} (x,y) =  \\min _{(x',y'):  \\, \\texttt{element} (x',y') \\ne0 } \\texttt{src} (x+x',y+y')\\f]\n",
      "    .   \n",
      "    .   The function supports the in-place mode. Erosion can be applied several ( iterations ) times. In\n",
      "    .   case of multi-channel images, each channel is processed independently.\n",
      "    .   \n",
      "    .   @param src input image; the number of channels can be arbitrary, but the depth should be one of\n",
      "    .   CV_8U, CV_16U, CV_16S, CV_32F or CV_64F.\n",
      "    .   @param dst output image of the same size and type as src.\n",
      "    .   @param kernel structuring element used for erosion; if `element=Mat()`, a `3 x 3` rectangular\n",
      "    .   structuring element is used. Kernel can be created using #getStructuringElement.\n",
      "    .   @param anchor position of the anchor within the element; default value (-1, -1) means that the\n",
      "    .   anchor is at the element center.\n",
      "    .   @param iterations number of times erosion is applied.\n",
      "    .   @param borderType pixel extrapolation method, see #BorderTypes\n",
      "    .   @param borderValue border value in case of a constant border\n",
      "    .   @sa  dilate, morphologyEx, getStructuringElement\n",
      "\n"
     ]
    }
   ],
   "source": [
    "help(cv2.erode)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "-1"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "img_temp1 = cv2.erode(thres, vertical_kernel, iterations=3)\n",
    "verticle_lines = cv2.dilate(img_temp1, vertical_kernel, iterations=3)\n",
    "\n",
    "cv2.imshow('verticle',verticle_lines)\n",
    "cv2.waitKey(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "-1"
      ]
     },
     "execution_count": 77,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "img_temp2 = cv2.erode(thres, horizontal_kernel, iterations=2)\n",
    "horizontal_lines = cv2.dilate(img_temp2, horizontal_kernel, iterations=20)\n",
    "\n",
    "cv2.imshow('horizontal',horizontal_lines)\n",
    "cv2.waitKey(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "-1"
      ]
     },
     "execution_count": 80,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "alpha = 0.5\n",
    "beta = 1-alpha\n",
    "\n",
    "img_final_bin = cv2.addWeighted(verticle_lines, alpha, horizontal_lines, beta,0.0)\n",
    "img_final_bin = cv2.erode(~img_final_bin, kernel, iterations=2)\n",
    "(thresh, img_final_bin) = cv2.threshold(img_final_bin, 128,  255,\n",
    "                                        cv2.THRESH_BINARY+cv2.THRESH_OTSU)\n",
    "\n",
    "cv2.imshow('wat up',img_final_bin)\n",
    "cv2.waitKey(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Help on built-in function addWeighted:\n",
      "\n",
      "addWeighted(...)\n",
      "    addWeighted(src1, alpha, src2, beta, gamma[, dst[, dtype]]) -> dst\n",
      "    .   @brief Calculates the weighted sum of two arrays.\n",
      "    .   \n",
      "    .   The function addWeighted calculates the weighted sum of two arrays as follows:\n",
      "    .   \\f[\\texttt{dst} (I)= \\texttt{saturate} ( \\texttt{src1} (I)* \\texttt{alpha} +  \\texttt{src2} (I)* \\texttt{beta} +  \\texttt{gamma} )\\f]\n",
      "    .   where I is a multi-dimensional index of array elements. In case of multi-channel arrays, each\n",
      "    .   channel is processed independently.\n",
      "    .   The function can be replaced with a matrix expression:\n",
      "    .   @code{.cpp}\n",
      "    .   dst = src1*alpha + src2*beta + gamma;\n",
      "    .   @endcode\n",
      "    .   @note Saturation is not applied when the output array has the depth CV_32S. You may even get\n",
      "    .   result of an incorrect sign in the case of overflow.\n",
      "    .   @param src1 first input array.\n",
      "    .   @param alpha weight of the first array elements.\n",
      "    .   @param src2 second input array of the same size and channel number as src1.\n",
      "    .   @param beta weight of the second array elements.\n",
      "    .   @param gamma scalar added to each sum.\n",
      "    .   @param dst output array that has the same size and number of channels as the input arrays.\n",
      "    .   @param dtype optional depth of the output array; when both input arrays have the same depth, dtype\n",
      "    .   can be set to -1, which will be equivalent to src1.depth().\n",
      "    .   @sa  add, subtract, scaleAdd, Mat::convertTo\n",
      "\n"
     ]
    }
   ],
   "source": [
    "help(cv2.addWeighted)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [
    {
     "ename": "ModuleNotFoundError",
     "evalue": "No module named 'pyimagesearch'",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mModuleNotFoundError\u001b[0m                       Traceback (most recent call last)",
      "\u001b[1;32m<ipython-input-81-4e2b668791bf>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m\u001b[0m\n\u001b[1;32m----> 1\u001b[1;33m \u001b[1;32mfrom\u001b[0m \u001b[0mpyimagesearch\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mtransform\u001b[0m \u001b[1;32mimport\u001b[0m \u001b[0mfour_point_transform\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[1;31mModuleNotFoundError\u001b[0m: No module named 'pyimagesearch'"
     ]
    }
   ],
   "source": [
    "from pyimagesearch.transform import four_point_transform"
   ]
  },
  {
   "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": "3.7.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}