Random_shapes.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Generating random shapes\n",
    "### Dr. Tirthajyoti Sarkar, Fremont, CA 94536\n",
    "\n",
    "This notebook demonstrates generation of random shapes with various properties.<br>\n",
    "It uses the `random_shapes()` function from `skimage.draw`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from skimage.draw import random_shapes"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Let's start simple and generate a 128x128 image with a single grayscale rectangle.\n",
    "We get back a tuple consisting of \n",
    "1. The image with the generated shapes and \n",
    "2. A list of label tuples with the kind of shape (e.g. circle, rectangle) and ((r0, r1), (c0, c1)) coordinates."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "rect = random_shapes((128, 128), max_shapes=1, shape='rectangle',\n",
    "                       multichannel=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Image shape: (128, 128)\n",
      "Labels: [('rectangle', ((45, 69), (20, 52)))]\n"
     ]
    }
   ],
   "source": [
    "image, labels = rect\n",
    "print('Image shape: {}\\nLabels: {}'.format(image.shape, labels))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 288x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(nrows=1, ncols=1,figsize=(4,4))\n",
    "ax.imshow(image, cmap='gray')\n",
    "ax.set_title('Grayscale shape')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### The generated images can be much more complex \n",
    "For example, let's try many shapes of any color. If we want the colors to be particularly light, we can set the `intensity_range` to an upper subrange of (0,255)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "image1, _ = random_shapes((128, 128), max_shapes=10,\n",
    "                          intensity_range=((100, 255),))\n",
    "image2, _ = random_shapes((128, 128), max_shapes=10,\n",
    "                          intensity_range=((200, 255),))\n",
    "image3, _ = random_shapes((128, 128), max_shapes=10,\n",
    "                          intensity_range=((50, 255),))\n",
    "image4, _ = random_shapes((128, 128), max_shapes=10,\n",
    "                          intensity_range=((0, 255),))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x576 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(nrows=2, ncols=2,figsize=(8,8))\n",
    "ax = axes.ravel()\n",
    "for i, image in enumerate([image1, image2, image3, image4], 0):\n",
    "    ax[i].imshow(image)\n",
    "    ax[i].set_title('Colored shapes, #{}'.format(i-1))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Overlapping shapes\n",
    "These shapes are well suited to test segmentation algorithms. Often, we want shapes to overlap to test the algorithm."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "image, _ = random_shapes((128, 128), min_shapes=5, max_shapes=6,\n",
    "                         min_size=20, allow_overlap=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(nrows=1, ncols=1,figsize=(5,5))\n",
    "ax.imshow(image)\n",
    "ax.set_title('Overlapping shapes')\n",
    "ax.set_xticklabels([])\n",
    "ax.set_yticklabels([])\n",
    "\n",
    "plt.show()"
   ]
  }
 ],
 "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": "3.6.2"
  },
  "latex_envs": {
   "LaTeX_envs_menu_present": true,
   "autoclose": false,
   "autocomplete": true,
   "bibliofile": "biblio.bib",
   "cite_by": "apalike",
   "current_citInitial": 1,
   "eqLabelWithNumbers": true,
   "eqNumInitial": 1,
   "hotkeys": {
    "equation": "Ctrl-E",
    "itemize": "Ctrl-I"
   },
   "labels_anchors": false,
   "latex_user_defs": false,
   "report_style_numbering": false,
   "user_envs_cfg": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}