ipcv/python_notebook/hough.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import imageio\n",
    "import math"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "def rgb2gray(rgb):\n",
    "    return np.dot(rgb[..., :3], [0.299, 0.587, 0.114]).astype(np.uint8)\n",
    "\n",
    "\n",
    "def hough_line(img, angle_step=1, lines_are_white=True, value_threshold=5):\n",
    "    \"\"\"\n",
    "    Hough transform for lines\n",
    "\n",
    "    Input:\n",
    "    img - 2D binary image with nonzeros representing edges\n",
    "    angle_step - Spacing between angles to use every n-th angle\n",
    "                 between -90 and 90 degrees. Default step is 1.\n",
    "    lines_are_white - boolean indicating whether lines to be detected are white\n",
    "    value_threshold - Pixel values above or below the value_threshold are edges\n",
    "\n",
    "    Returns:\n",
    "    accumulator - 2D array of the hough transform accumulator\n",
    "    theta - array of angles used in computation, in radians.\n",
    "    rhos - array of rho values. Max size is 2 times the diagonal\n",
    "           distance of the input image.\n",
    "    \"\"\"\n",
    "    # Rho and Theta ranges\n",
    "    thetas = np.deg2rad(np.arange(-90.0, 90.0, angle_step))\n",
    "    width, height = img.shape\n",
    "    diag_len = int(round(math.sqrt(width * width + height * height)))\n",
    "    rhos = np.linspace(-diag_len, diag_len, diag_len * 2)\n",
    "\n",
    "    # Cache some resuable values\n",
    "    cos_t = np.cos(thetas)\n",
    "    sin_t = np.sin(thetas)\n",
    "    num_thetas = len(thetas)\n",
    "\n",
    "    # Hough accumulator array of theta vs rho\n",
    "    accumulator = np.zeros((2 * diag_len, num_thetas), dtype=np.uint8)\n",
    "    # (row, col) indexes to edges\n",
    "    are_edges = img > value_threshold if lines_are_white else img < value_threshold\n",
    "    y_idxs, x_idxs = np.nonzero(are_edges)\n",
    "\n",
    "    # Vote in the hough accumulator\n",
    "    for i in range(len(x_idxs)):\n",
    "        x = x_idxs[i]\n",
    "        y = y_idxs[i]\n",
    "\n",
    "        for t_idx in range(num_thetas):\n",
    "            # Calculate rho. diag_len is added for a positive index\n",
    "            rho = diag_len + int(round(x * cos_t[t_idx] + y * sin_t[t_idx]))\n",
    "            accumulator[rho, t_idx] += 1\n",
    "\n",
    "    return accumulator, thetas, rhos"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "def show_hough_line(img, accumulator, thetas, rhos, save_path=None):\n",
    "    import matplotlib.pyplot as plt\n",
    "\n",
    "    fig, ax = plt.subplots(1, 2, figsize=(10, 10))\n",
    "\n",
    "    ax[0].imshow(img, cmap=plt.cm.gray)\n",
    "    ax[0].set_title('Input image')\n",
    "    ax[0].axis('image')\n",
    "\n",
    "    ax[1].imshow(\n",
    "        accumulator, cmap='jet',\n",
    "        extent=[np.rad2deg(thetas[-1]), np.rad2deg(thetas[0]), rhos[-1], rhos[0]])\n",
    "    ax[1].set_aspect('equal', adjustable='box')\n",
    "    ax[1].set_title('Hough transform')\n",
    "    ax[1].set_xlabel('Angles (degrees)')\n",
    "    ax[1].set_ylabel('Distance (pixels)')\n",
    "    ax[1].axis('image')\n",
    "\n",
    "    # plt.axis('off')\n",
    "    if save_path is not None:\n",
    "        plt.savefig(save_path, bbox_inches='tight')\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "imgpath = '../images/bangdiem.png'\n",
    "img = imageio.imread(imgpath)\n",
    "if img.ndim == 3:\n",
    "    img = rgb2gray(img)\n",
    "accumulator, thetas, rhos = hough_line(img)"
   ]
  }
 ],
 "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": "3.7.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}