change_label.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import glob\n",
    "import random\n",
    "import json\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "from pathlib import Path\n",
    "from tqdm.auto import tqdm,trange\n",
    "import csv\n",
    "from PIL import Image\n",
    "from PIL import ImageFile\n",
    "ImageFile.LOAD_TRUNCATED_IMAGES = True\n",
    "\n",
    "# import torch\n",
    "# import torchvision\n",
    "# from torchvision import transforms\n",
    "import matplotlib.pyplot as plt\n",
    "from pandas import DataFrame"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "30000 30000\n",
      "30000 30000\n"
     ]
    }
   ],
   "source": [
    "def number_of_subfiles(train_dir):\n",
    "    s = 0\n",
    "    lists = []\n",
    "    for (path, dir, files) in os.walk(train_dir):\n",
    "        for filename in files:\n",
    "            ext = os.path.splitext(filename)[-1]\n",
    "            # print(path+\"/\"+filename)\n",
    "            if ext != '.zip':\n",
    "\n",
    "                file_index = os.path.splitext(filename)[0]\n",
    "                if int(file_index[-6:]) >= 30000 and int(file_index[-6:]) < 60000:\n",
    "                    s+=1\n",
    "                    # print(os.path.splitext(filename)[0])\n",
    "                    lists.append(path+\"/\"+filename)\n",
    "    return s, lists\n",
    "\n",
    "# cnt_image, list_image = number_of_subfiles(\"/hdd1/image/1. 가로형간판/\")\n",
    "# cnt_label, list_label = number_of_subfiles(\"/hdd1/label/1.간판/1.가로형간판/가로형간판2/\")\n",
    "cnt_image, list_image = number_of_subfiles(\"D:\\Dacon\\야외 실제 촬영 한글 이미지\\[원천]Training_간판_가로형간판_원천데이터2\")\n",
    "# D:\\Dacon\\야외 실제 촬영 한글 이미지\\[라벨]Training\\1.간판\\1.가로형간판\\가로형간판2\n",
    "cnt_label, list_label = number_of_subfiles(\"D:/Dacon/야외 실제 촬영 한글 이미지/[라벨]Training/1.간판/1.가로형간판/가로형간판2\")\n",
    "\n",
    "print(cnt_image, list_image.__len__())\n",
    "print(cnt_label, list_label.__len__())\n",
    "# print(list_image[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "list_image = sorted(list_image)\n",
    "list_label = sorted(list_label)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "('D:\\\\Dacon\\\\야외 실제 촬영 한글 이미지\\\\[원천]Training_간판_가로형간판_원천데이터2/간판_가로형간판_030000.jpg',\n",
       " 'D:/Dacon/야외 실제 촬영 한글 이미지/[라벨]Training/1.간판/1.가로형간판/가로형간판2/간판_가로형간판_030000.json')"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "list_image[0], list_label[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e23413c508fa4af7b2ca390d8d6fb0ba",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/30000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "json_list = []\n",
    "for label in tqdm(list_label):    \n",
    "    with open(label, encoding='utf8') as f:\n",
    "        json_list.append(json.load(f))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'annotations': [{'bbox': [435, 366, 937, 431],\n",
       "   'id': 1,\n",
       "   'text': 'xxx',\n",
       "   'image_id': 1},\n",
       "  {'bbox': [141, 482, 259, 419], 'id': 2, 'text': '늘푸른', 'image_id': 1}],\n",
       " 'cropLabels': [],\n",
       " 'images': [{'file_name': '간판_가로형간판_030000.jpg',\n",
       "   'date_created': '20210126',\n",
       "   'width': 1600,\n",
       "   'id': 1,\n",
       "   'height': 1200}],\n",
       " 'info': {'date_created': '20210126',\n",
       "  'name': '간판_가로형간판_030000',\n",
       "  'description': ''},\n",
       " 'metadata': [{'area': '경상권',\n",
       "   'wordfont': '캘리그라피',\n",
       "   'outline': '상',\n",
       "   'illuminance': '밝음',\n",
       "   'wordsize': '중',\n",
       "   'light': '자연광',\n",
       "   'wordconnection': '빈공간없이 이어짐',\n",
       "   'subclass': '가로형간판',\n",
       "   'weather': '맑음',\n",
       "   'device': '스마트폰',\n",
       "   'wordcolor': '단색',\n",
       "   'class': '실외간판',\n",
       "   'wordorientation': '세로'}]}"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "json_list[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b7d605d104464011b274c0615dbdd4c2",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/30000 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "label = []\n",
    "lotate  = []\n",
    "idxs = []\n",
    "# list_image, list_label\n",
    "\n",
    "# idx -> number, i is info\n",
    "for idx,i in tqdm(enumerate(json_list),total=len(json_list)):\n",
    "    for jdx,j in enumerate(i['annotations']):\n",
    "        # if j['bbox'] == [None, None, None, None]: continue\n",
    "        # if j['bbox'] == [None, None, None, None]: \n",
    "        if j['text'] == \"xxx\": \n",
    "            idxs.append(idx)\n",
    "            lotate.append(j['bbox'])\n",
    "            label.append(j['text'])\n",
    "        # elif j['text'] == \"xxx\": continue\n",
    "        # elif j['text'] == \" \": continue\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(69901, 69901, 69901)"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "idxs.__len__(), lotate.__len__(), label.__len__()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f44b9098c4974c2491237f73e70d1a9b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/69901 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "for i in trange(len(lotate)):\n",
    "    lotate[i][2] = lotate[i][0]+lotate[i][2]\n",
    "    lotate[i][3] = lotate[i][1]+lotate[i][3]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[[435, 366, 1372, 797],\n",
       " [368, 0, 1152, 299],\n",
       " [0, 384, 613, 497],\n",
       " [0, 714, 274, 798],\n",
       " [1063, 1148, 1347, 1200]]"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lotate[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "9136a6b3673c465190fbe2a4cbb0bfa4",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/58390 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "ename": "SystemError",
     "evalue": "tile cannot extend outside image",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mSystemError\u001b[0m                               Traceback (most recent call last)",
      "\u001b[1;32m~\\AppData\\Local\\Temp/ipykernel_24692/475392922.py\u001b[0m in \u001b[0;36m<module>\u001b[1;34m\u001b[0m\n\u001b[0;32m     12\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     13\u001b[0m     \u001b[1;31m# crop_image.resize((crop_image.size[0]*alpha,crop_image.size[1]*alpha))#resize하고 싶은 비율 설정\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 14\u001b[1;33m     \u001b[0mcrop_image\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0msave\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;34mf'{path}/{list_image[idxs[idx]].split(\".\")[0][-6:]} and {i} train_data_{idx}.jpg'\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[1;32m~\\anaconda3\\lib\\site-packages\\PIL\\Image.py\u001b[0m in \u001b[0;36msave\u001b[1;34m(self, fp, format, **params)\u001b[0m\n\u001b[0;32m   2238\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   2239\u001b[0m         \u001b[1;32mtry\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m-> 2240\u001b[1;33m             \u001b[0msave_handler\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mfp\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mfilename\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m   2241\u001b[0m         \u001b[1;32mfinally\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   2242\u001b[0m             \u001b[1;31m# do what we can to clean up\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\anaconda3\\lib\\site-packages\\PIL\\JpegImagePlugin.py\u001b[0m in \u001b[0;36m_save\u001b[1;34m(im, fp, filename)\u001b[0m\n\u001b[0;32m    780\u001b[0m     \u001b[0mbufsize\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mmax\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mImageFile\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mMAXBLOCK\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mbufsize\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mlen\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mexif\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;33m+\u001b[0m \u001b[1;36m5\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mlen\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mextra\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;33m+\u001b[0m \u001b[1;36m1\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    781\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 782\u001b[1;33m     \u001b[0mImageFile\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_save\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mim\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mfp\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m[\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;34m\"jpeg\"\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m(\u001b[0m\u001b[1;36m0\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;36m0\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;33m+\u001b[0m \u001b[0mim\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0msize\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;36m0\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mrawmode\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mbufsize\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    783\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    784\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\anaconda3\\lib\\site-packages\\PIL\\ImageFile.py\u001b[0m in \u001b[0;36m_save\u001b[1;34m(im, fp, tile, bufsize)\u001b[0m\n\u001b[0;32m    529\u001b[0m             \u001b[1;32mif\u001b[0m \u001b[0mo\u001b[0m \u001b[1;33m>\u001b[0m \u001b[1;36m0\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    530\u001b[0m                 \u001b[0mfp\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mseek\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mo\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 531\u001b[1;33m             \u001b[0me\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0msetimage\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mim\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mim\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mb\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    532\u001b[0m             \u001b[1;32mif\u001b[0m \u001b[0me\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mpushes_fd\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    533\u001b[0m                 \u001b[0me\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0msetfd\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mfp\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;31mSystemError\u001b[0m: tile cannot extend outside image"
     ]
    }
   ],
   "source": [
    "new_image = []\n",
    "\n",
    "path = \"D:/Dacon/crop_image/\"\n",
    "# alpha = 0.5 #resize할 범위 만약 할거면 값 조정해서 resize해서 저장하면 됨 \n",
    "start = 0 # default\n",
    "starts = 11510 # 하다가 오류 떴을 때, 뜬 번호 여기에 넣고 하면 이어서 가능\n",
    "# bins = 1\n",
    "for idx,i in tqdm(enumerate(lotate[starts:]),total =len(lotate[starts:])):\n",
    "    idx+=starts\n",
    "    image = Image.open(list_image[idxs[idx]])\n",
    "    crop_image = image.crop(i)\n",
    "\n",
    "    # crop_image.resize((crop_image.size[0]*alpha,crop_image.size[1]*alpha))#resize하고 싶은 비율 설정 \n",
    "    crop_image.save(f'{path}/{list_image[idxs[idx]].split(\".\")[0][-6:]} and {i} train_data_{idx}.jpg')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 만약 오류가 떴다면"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [],
   "source": [
    "# lotate[11510+69:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1600, 1200)"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Image.open(list_image[idxs[idx]]) # 이미지 확인\n",
    "Image.open(list_image[idxs[11510+69]]).size # 사이즈 확인"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "([1592, 1206, 1600], 4874, 'xxx')"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## 디버깅 귀찮다면\n",
    "# 해당위치 = 11510\n",
    "# lotate.pop(해당위치), idxs.pop(해당위치), label.pop(해당위치),"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'imageScreenshot' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[1;32m~\\AppData\\Local\\Temp/ipykernel_24692/1652915376.py\u001b[0m in \u001b[0;36m<module>\u001b[1;34m\u001b[0m\n\u001b[1;32m----> 1\u001b[1;33m \u001b[0mimageScreenshot\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcrop\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mx\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0my\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mwidth\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mheight\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;31m# 으로 확인 가능\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[1;31mNameError\u001b[0m: name 'imageScreenshot' is not defined"
     ]
    }
   ],
   "source": [
    "imageScreenshot.crop((x, y, width, height)) # 으로 확인 가능"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.7"
  },
  "vscode": {
   "interpreter": {
    "hash": "02de751381aea847d986eeaeffcf423627f382a59ad69a145cf71f2f5424c15d"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}