sitelen-palisa.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "from PIL import Image, ImageDraw, ImageFilter\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "\n",
    "def flip_horizontal(l):\n",
    "    out = []\n",
    "    for i,j in l:\n",
    "        out.append((w-i,j))\n",
    "    return out\n",
    "\n",
    "def flip_vertical(l):\n",
    "    out = []\n",
    "    for i,j in l:\n",
    "        out.append((i,h-j))\n",
    "    return out\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {},
   "outputs": [],
   "source": [
    "w,h = 150,200"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {},
   "outputs": [],
   "source": [
    "vertical_margin = 20\n",
    "horizontal_margin = 20\n",
    "line_width = 10\n",
    "gap = 5\n",
    "tilt = 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "metadata": {},
   "outputs": [],
   "source": [
    "swid = line_width // 2\n",
    "topline = swid + vertical_margin\n",
    "leftline = swid + horizontal_margin\n",
    "center_vert = w/2\n",
    "center_hori = h/2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {},
   "outputs": [],
   "source": [
    "seg_coords = {\n",
    "    \"t\": [(30,25), (35,20), (115,20), (120,25), (115,30), (35,30)],\n",
    "    \"tl\":[(25,30), (20,35), (20,90), (25,95), (30,90), (30,35)],\n",
    "    \"ml\":[(30,100), (35,105), (65,105), (70,100), (65,95), (35,95)],\n",
    "    \"tm\":[(70,35), (80,35), (80,90), (75,95), (70,90)],\n",
    "    \"tld\":[(35,35), (35,50), (65,90), (65,75)]\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {},
   "outputs": [],
   "source": [
    "seg_coords[\"b\"] = flip_vertical(seg_coords[\"t\"])\n",
    "seg_coords[\"tr\"] = flip_horizontal(seg_coords[\"tl\"])\n",
    "seg_coords[\"bl\"] = flip_vertical(seg_coords[\"tl\"])\n",
    "seg_coords[\"br\"] = flip_horizontal(seg_coords[\"bl\"])\n",
    "seg_coords[\"mr\"] = flip_horizontal(seg_coords[\"ml\"])\n",
    "seg_coords[\"bm\"] = flip_vertical(seg_coords[\"tm\"])\n",
    "seg_coords[\"trd\"] = flip_horizontal(seg_coords[\"tld\"])\n",
    "seg_coords[\"bld\"] = flip_vertical(seg_coords[\"tld\"])\n",
    "seg_coords[\"brd\"] = flip_vertical(seg_coords[\"trd\"])"
   ]
  },
  {
   "cell_type": "raw",
   "metadata": {},
   "source": [
    "new_seg_coords = dict()\n",
    "seg_coords = {t:[(i-(j-h/2)*tilt, j) for i,j in poly] for t,poly in seg_coords.items()}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "metadata": {},
   "outputs": [],
   "source": [
    "lit_segs = [i for i in seg_coords.keys() if np.random.random()<0.5]\n",
    "lit_segs = [\"b\",\"tl\",\"tr\",\"ml\",\"mr\",\"bl\",\"br\",\"bm\",\"tld\",\"trd\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {},
   "outputs": [],
   "source": [
    "on_seg = Image.new(\"RGBA\",(w,h),(0,0,0,0))\n",
    "off_seg = Image.new(\"RGBA\",(w,h),(0,0,0,0))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {},
   "outputs": [],
   "source": [
    "on_seg_d = ImageDraw.Draw(on_seg)\n",
    "off_seg_d = ImageDraw.Draw(off_seg)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "metadata": {},
   "outputs": [],
   "source": [
    "k=np.zeros([w,h,4])\n",
    "\n",
    "hole_size = 4\n",
    "binary_mask = [[((y%(2*hole_size)==0) or ((x+(hole_size)*((y//(2*hole_size))%2))%(2*hole_size)==0)) for x in range(w)] for y in range(h)]\n",
    "\n",
    "Image.fromarray(255-np.array(binary_mask, dtype=\"uint8\")*255);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "metadata": {},
   "outputs": [],
   "source": [
    "for seg_name, seg_poly in seg_coords.items(): \n",
    "    if seg_name in lit_segs:\n",
    "        on_seg_d.polygon(seg_poly, fill=(0,255,255,255))\n",
    "    else:\n",
    "        off_seg_d.polygon(seg_poly, fill=(0,255,255,30))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<PIL.ImageDraw.ImageDraw at 0x7f94f7ca2310>"
      ]
     },
     "execution_count": 117,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "on_seg_d"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "metadata": {},
   "outputs": [],
   "source": [
    "out_image = Image.new(mode=\"RGBA\",size=(w,h), color=(0,20,20,255))\n",
    "out_image = Image.alpha_composite(out_image, on_seg)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "metadata": {},
   "outputs": [],
   "source": [
    "out_image = out_image.filter(ImageFilter.GaussianBlur(7)) #apply glow"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "metadata": {},
   "outputs": [],
   "source": [
    "out_image = Image.alpha_composite(out_image, on_seg)\n",
    "out_image = Image.alpha_composite(out_image, off_seg)\n",
    "out_image;"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {},
   "outputs": [],
   "source": [
    "grid_opacity = 50\n",
    "z=np.array(binary_mask)[:,:,np.newaxis]*(np.ones([1,1,4]))*grid_opacity\n",
    "z[:,:,:3]=0\n",
    "grid = Image.fromarray(z.astype(\"uint8\"))\n",
    "grid = grid.filter(ImageFilter.GaussianBlur(0.5))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "metadata": {},
   "outputs": [],
   "source": [
    "z[:,:,:3]=0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<PIL.Image.Image image mode=RGBA size=150x200 at 0x7F94F8305520>"
      ]
     },
     "execution_count": 123,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "out_image = Image.alpha_composite(out_image, grid)\n",
    "out_image"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'t': [(30, 25), (35, 20), (115, 20), (120, 25), (115, 30), (35, 30)],\n",
       " 'tl': [(25, 30), (20, 35), (20, 90), (25, 95), (30, 90), (30, 35)],\n",
       " 'ml': [(30, 100), (35, 105), (65, 105), (70, 100), (65, 95), (35, 95)],\n",
       " 'tm': [(70, 35), (80, 35), (80, 90), (75, 95), (70, 90)],\n",
       " 'tld': [(35, 35), (35, 50), (65, 90), (65, 75)],\n",
       " 'b': [(30, 175), (35, 180), (115, 180), (120, 175), (115, 170), (35, 170)],\n",
       " 'tr': [(125, 30), (130, 35), (130, 90), (125, 95), (120, 90), (120, 35)],\n",
       " 'bl': [(25, 170), (20, 165), (20, 110), (25, 105), (30, 110), (30, 165)],\n",
       " 'br': [(125, 170),\n",
       "  (130, 165),\n",
       "  (130, 110),\n",
       "  (125, 105),\n",
       "  (120, 110),\n",
       "  (120, 165)],\n",
       " 'mr': [(120, 100), (115, 105), (85, 105), (80, 100), (85, 95), (115, 95)],\n",
       " 'bm': [(70, 165), (80, 165), (80, 110), (75, 105), (70, 110)],\n",
       " 'trd': [(115, 35), (115, 50), (85, 90), (85, 75)],\n",
       " 'bld': [(35, 165), (35, 150), (65, 110), (65, 125)],\n",
       " 'brd': [(115, 165), (115, 150), (85, 110), (85, 125)]}"
      ]
     },
     "execution_count": 124,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "seg_coords"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "metadata": {},
   "outputs": [],
   "source": [
    "from json import dump"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 126,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open(\"seg_coords.json\",\"w\") as f:\n",
    "    dump(seg_coords, f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "metadata": {},
   "outputs": [],
   "source": [
    "from collections import defaultdict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "defaultdict(list,\n",
       "            {'t': [(30, 25),\n",
       "              (35, 20),\n",
       "              (115, 20),\n",
       "              (120, 25),\n",
       "              (115, 30),\n",
       "              (35, 30)],\n",
       "             'tl': [(25, 30),\n",
       "              (20, 35),\n",
       "              (20, 90),\n",
       "              (25, 95),\n",
       "              (30, 90),\n",
       "              (30, 35)],\n",
       "             'ml': [(30, 100),\n",
       "              (35, 105),\n",
       "              (65, 105),\n",
       "              (70, 100),\n",
       "              (65, 95),\n",
       "              (35, 95)],\n",
       "             'tm': [(70, 35), (80, 35), (80, 90), (75, 95), (70, 90)],\n",
       "             'tld': [(35, 35), (35, 50), (65, 90), (65, 75)],\n",
       "             'b': [(30, 175),\n",
       "              (35, 180),\n",
       "              (115, 180),\n",
       "              (120, 175),\n",
       "              (115, 170),\n",
       "              (35, 170)],\n",
       "             'tr': [(125, 30),\n",
       "              (130, 35),\n",
       "              (130, 90),\n",
       "              (125, 95),\n",
       "              (120, 90),\n",
       "              (120, 35)],\n",
       "             'bl': [(25, 170),\n",
       "              (20, 165),\n",
       "              (20, 110),\n",
       "              (25, 105),\n",
       "              (30, 110),\n",
       "              (30, 165)],\n",
       "             'br': [(125, 170),\n",
       "              (130, 165),\n",
       "              (130, 110),\n",
       "              (125, 105),\n",
       "              (120, 110),\n",
       "              (120, 165)],\n",
       "             'mr': [(120, 100),\n",
       "              (115, 105),\n",
       "              (85, 105),\n",
       "              (80, 100),\n",
       "              (85, 95),\n",
       "              (115, 95)],\n",
       "             'bm': [(70, 165), (80, 165), (80, 110), (75, 105), (70, 110)],\n",
       "             'trd': [(115, 35), (115, 50), (85, 90), (85, 75)],\n",
       "             'bld': [(35, 165), (35, 150), (65, 110), (65, 125)],\n",
       "             'brd': [(115, 165), (115, 150), (85, 110), (85, 125)]})"
      ]
     },
     "execution_count": 129,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "defaultdict(list, seg_coords)"
   ]
  },
  {
   "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": "3.8.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}