PastePicture2Wallpaper.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "SwUzOv_YJ5c_"
      },
      "source": [
        "# PastePicture2Wallpaper　\n",
        "\n",
        "壁紙作成のための線画作成ツールです🎠\n",
        "\n",
        "**GPUランタイムで実行してください**\n",
        "\n",
        "<br>\n",
        "\n",
        "処理手順\n",
        "1. 深度閾値に従ってMidasを用いた単眼深度推定を行います\n",
        "2. Canny法による線画抽出を行います(線画化過程でノイズ除去や線を太くするなどの処理を行っています)\n",
        "3. OSの雰囲気に応じて半透明化した線画を出力します（result.pngという名前で出力されます）\n",
        "\n",
        "<br>\n",
        "\n",
        "🆕 : informative_drawingsが有効になりました。（デフォルト：True）\n",
        "\n",
        "informative_drawingsはAIによって線画抽出を行います。\n",
        "\n",
        "- median_k_sizeとmodeのオプションは無視されます。\n",
        "\n",
        "<br>\n",
        "\n",
        "以下のパラメータで値を弄ってください\n",
        "\n",
        "\n",
        "<!-- - image_upload_option: 画像をアップロードする方法 -->\n",
        "  <!-- - `Upload from device`: デバイスからアップロード -->\n",
        "  <!-- - `URL`: URLからダウンロード -->\n",
        "  <!-- - `Mount Drive`: Google Driveのマウントを行う -->\n",
        "<!-- - path_or_url: ダウンロード先のURLかマウント先のデバイスのパスを指定(Upload from devieの場合は無視されます) -->\n",
        "- mode: 線画処理のモード\n",
        "  - Anime : アニメ風の輪郭がはっきりとしている画像向け\n",
        "- OS: 壁紙のOSの種類によって出力される線画の雰囲気が少し変わります\n",
        "  - Mac風(灰色):パステルカラーの背景に合わせて半透明の灰色線画を出力\n",
        "  - Mac風(白色):鮮やかな背景に半透明の白線画を出力\n",
        "  - Windows風: 水色の背景にさらに明るい水色半透明の線画を出力\n",
        "  - Ubuntu風: 紫-桃色の背景に馴染む桃色半透明の線画を出力\n",
        "- median_k_size: 細かなノイズを除去します。1の場合、無効化します。\n",
        "- morph_g_ksize: モロフォジー変換用カーネルサイズ。膨張した画像と収縮した画像の差分をとる処理を行い、物体の輪郭を得ます。1の場合は無効化します。\n",
        "- depth_threshold: 出力された深度情報をもとに線画化するための閾値\n",
        "\n",
        "\n",
        "\n",
        "![](https://raw.githubusercontent.com/Ar-Ray-code/PastePicture2Wallpaper/main/images_for_readme/export-variation.png)\n",
        "\n",
        "> デバイスからのアップロードのみ受け付けています\n",
        "> \n",
        "> 直接URLを指定する場合は4番目のコードセルのURLを変更してコメントアウトを外し、以降のセルを実行（Ctrl+F10） を行ってください"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {
        "id": "9UIXNyyXdKaO"
      },
      "outputs": [],
      "source": [
        "# image_upload_option = 'Upload from device' #@param [\"Upload from device\", \"URL\", \"Mount Drive\"] {allow-input: true}\n",
        "# path_or_url = \"/content/sample.jpg\"#@param {type:\"string\"}\n",
        "\n",
        "mode = 'Anime' #@param [\"Anime\"] {allow-input: true}\n",
        "OS =  'Mac-Gray' #@param [\"Ubuntu\",\"Windows\",\"Mac-White\",\"Mac-Gray\"] {allow-input: true}\n",
        "using_informative_drawings = \"True\" #@param [\"True\", \"False\"] {allow-input: true}\n",
        "remove_back = 'ISNet' #@param [\"ISNet\", \"Midas\"]\n",
        "\n",
        "# median_ksize=median_ksize, morph_g_ksize=morph_g_ksize\n",
        "\n",
        "median_ksize = 5 #@param {type:\"slider\", min:1, max:9, step:2}\n",
        "morph_g_ksize = 5 #@param {type:\"slider\", min:1, max:9, step:2}\n",
        "depth_threshold = 150 #@param {type:\"slider\", min:0, max:255, step:1}\n",
        "canny_threshold = 150 #@param {type:\"slider\", min:0, max:255, step:1}\n",
        "\n",
        "default_picture = \"True\" #@param [\"True\", \"False\"] {allow-input: true}"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print(\"mode=\",mode)\n",
        "print(\"OS=\",OS)\n",
        "print(\"using_informative_drawings=\",using_informative_drawings)\n",
        "print(\"remove_back=\",remove_back)\n",
        "print(\"median_ksize=\",median_ksize)\n",
        "print(\"morph_g_ksize=\",morph_g_ksize)\n",
        "print(\"depth_threshold=\",depth_threshold)\n",
        "print(\"canny_threshold=\",canny_threshold)\n",
        "print(\"default_picture=\",default_picture)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "hXEDMxIxs7sT",
        "outputId": "bb761d78-b337-41e8-8a6e-75eb3cde65f0"
      },
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "mode= Anime\n",
            "OS= Mac-Gray\n",
            "using_informative_drawings= True\n",
            "remove_back= ISNet\n",
            "median_ksize= 5\n",
            "morph_g_ksize= 5\n",
            "depth_threshold= 150\n",
            "canny_threshold= 150\n",
            "default_picture= True\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "!pip install wget"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "jaFyU0CnwWz8",
        "outputId": "40e54975-d424-46ab-84bd-7ba3c23f5c27"
      },
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n",
            "Collecting wget\n",
            "  Downloading wget-3.2.zip (10 kB)\n",
            "  Preparing metadata (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "Building wheels for collected packages: wget\n",
            "  Building wheel for wget (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "  Created wheel for wget: filename=wget-3.2-py3-none-any.whl size=9657 sha256=edc7fe659dfecde18ab938f67187e8d65e90b3f9c0c3a895a0f72cc42cabc8ee\n",
            "  Stored in directory: /root/.cache/pip/wheels/8b/f1/7f/5c94f0a7a505ca1c81cd1d9208ae2064675d97582078e6c769\n",
            "Successfully built wget\n",
            "Installing collected packages: wget\n",
            "Successfully installed wget-3.2\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "!cd /content/\n",
        "!rm /content/*.*g"
      ],
      "metadata": {
        "id": "alrXbJMnDigz",
        "outputId": "13c8b90a-fbcd-4e5f-8124-faf76b879a0e",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "rm: cannot remove '/content/*.*g': No such file or directory\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 5,
      "metadata": {
        "id": "8PbJjVsrTrzK"
      },
      "outputs": [],
      "source": [
        "# Upload from device\n",
        "import os\n",
        "if default_picture == \"True\":\n",
        "    import wget\n",
        "    os.chdir('/content/')\n",
        "    url = 'https://s3.ap-northeast-1.wasabisys.com/download-raw/urara/album/urara-toresen.jpg'\n",
        "    filename = wget.download(url)\n",
        "else:\n",
        "    from google.colab import files\n",
        "    uploaded = files.upload()\n",
        "\n",
        "#@markdown ### アップロードを (Choose Files)から行ってください\n",
        "#@markdown Driveのマウントの場合はこのステップを飛ばして、新しくコードセルを作成して、`/content`に対象ファイルをコピーしてください\n",
        "\n",
        "# from google.colab import drive\n",
        "# drive.mount('/content/drive')\n",
        "\n",
        "#@markdown <br>\n",
        "\n",
        "#@markdown ### default_picture=Trueの場合\n",
        "#@markdown - [urara-toresen.jpg](https://s3.ap-northeast-1.wasabisys.com/download-raw/urara/album/urara-toresen.jpg) がダウンロードされます。\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 6,
      "metadata": {
        "id": "maDiodVfAc1F"
      },
      "outputs": [],
      "source": [
        "#@markdown ###URLを指定する場合はこちらを編集し、upload.jpgで保存すること\n",
        "# !rm sample.jpg\n",
        "# !wget \"https://drive.google.com/uc?export=download&id=1M1eDdinWr-aPT28cyOo01LlMFUnsBbRQ\" -O upload.jpg"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 7,
      "metadata": {
        "id": "LofUpIu2f2bJ",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "58e1944e-aca5-44d3-82fe-463824b72469"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "ffmpeg version 4.2.7-0ubuntu0.1 Copyright (c) 2000-2022 the FFmpeg developers\n",
            "  built with gcc 9 (Ubuntu 9.4.0-1ubuntu1~20.04.1)\n",
            "  configuration: --prefix=/usr --extra-version=0ubuntu0.1 --toolchain=hardened --libdir=/usr/lib/x86_64-linux-gnu --incdir=/usr/include/x86_64-linux-gnu --arch=amd64 --enable-gpl --disable-stripping --enable-avresample --disable-filter=resample --enable-avisynth --enable-gnutls --enable-ladspa --enable-libaom --enable-libass --enable-libbluray --enable-libbs2b --enable-libcaca --enable-libcdio --enable-libcodec2 --enable-libflite --enable-libfontconfig --enable-libfreetype --enable-libfribidi --enable-libgme --enable-libgsm --enable-libjack --enable-libmp3lame --enable-libmysofa --enable-libopenjpeg --enable-libopenmpt --enable-libopus --enable-libpulse --enable-librsvg --enable-librubberband --enable-libshine --enable-libsnappy --enable-libsoxr --enable-libspeex --enable-libssh --enable-libtheora --enable-libtwolame --enable-libvidstab --enable-libvorbis --enable-libvpx --enable-libwavpack --enable-libwebp --enable-libx265 --enable-libxml2 --enable-libxvid --enable-libzmq --enable-libzvbi --enable-lv2 --enable-omx --enable-openal --enable-opencl --enable-opengl --enable-sdl2 --enable-libdc1394 --enable-libdrm --enable-libiec61883 --enable-nvenc --enable-chromaprint --enable-frei0r --enable-libx264 --enable-shared\n",
            "  libavutil      56. 31.100 / 56. 31.100\n",
            "  libavcodec     58. 54.100 / 58. 54.100\n",
            "  libavformat    58. 29.100 / 58. 29.100\n",
            "  libavdevice    58.  8.100 / 58.  8.100\n",
            "  libavfilter     7. 57.100 /  7. 57.100\n",
            "  libavresample   4.  0.  0 /  4.  0.  0\n",
            "  libswscale      5.  5.100 /  5.  5.100\n",
            "  libswresample   3.  5.100 /  3.  5.100\n",
            "  libpostproc    55.  5.100 / 55.  5.100\n",
            "Input #0, image2, from 'urara-toresen.jpg':\n",
            "  Duration: 00:00:00.04, start: 0.000000, bitrate: 518236 kb/s\n",
            "    Stream #0:0: Video: mjpeg (Baseline), yuvj444p(pc, bt470bg/unknown/unknown), 1640x2148 [SAR 144:144 DAR 410:537], 25 tbr, 25 tbn, 25 tbc\n",
            "Stream mapping:\n",
            "  Stream #0:0 -> #0:0 (mjpeg (native) -> mjpeg (native))\n",
            "Press [q] to stop, [?] for help\n",
            "\u001b[1;34m[swscaler @ 0x55abd3332940] \u001b[0m\u001b[0;33mdeprecated pixel format used, make sure you did set range correctly\n",
            "\u001b[0mOutput #0, image2, to 'upload.jpg':\n",
            "  Metadata:\n",
            "    encoder         : Lavf58.29.100\n",
            "    Stream #0:0: Video: mjpeg, yuvj444p(pc), 550x720 [SAR 1968:1969 DAR 410:537], q=2-31, 200 kb/s, 25 fps, 25 tbn, 25 tbc\n",
            "    Metadata:\n",
            "      encoder         : Lavc58.54.100 mjpeg\n",
            "    Side data:\n",
            "      cpb: bitrate max/min/avg: 0/0/200000 buffer size: 0 vbv_delay: -1\n",
            "frame=    1 fps=0.0 q=6.8 Lsize=N/A time=00:00:00.04 bitrate=N/A speed=0.174x    \n",
            "video:72kB audio:0kB subtitle:0kB other streams:0kB global headers:0kB muxing overhead: unknown\n"
          ]
        }
      ],
      "source": [
        "!ffmpeg -i *.*g -vf scale=-1:720 upload.jpg"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 8,
      "metadata": {
        "id": "Z81Rg6PoAhDc",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "fa86d820-762a-4ae1-d14a-67346d7879bd"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n",
            "Collecting tensorflow==2.8.0\n",
            "  Downloading tensorflow-2.8.0-cp310-cp310-manylinux2010_x86_64.whl (497.6 MB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m497.6/497.6 MB\u001b[0m \u001b[31m2.8 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hCollecting tensorflow-hub==0.12.0\n",
            "  Downloading tensorflow_hub-0.12.0-py2.py3-none-any.whl (108 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m108.8/108.8 kB\u001b[0m \u001b[31m12.1 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hCollecting onnxruntime==1.15.0\n",
            "  Downloading onnxruntime-1.15.0-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (5.9 MB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m5.9/5.9 MB\u001b[0m \u001b[31m51.9 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hCollecting onnx==1.14.0\n",
            "  Downloading onnx-1.14.0-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (14.6 MB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m14.6/14.6 MB\u001b[0m \u001b[31m67.0 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hRequirement already satisfied: absl-py>=0.4.0 in /usr/local/lib/python3.10/dist-packages (from tensorflow==2.8.0) (1.4.0)\n",
            "Requirement already satisfied: astunparse>=1.6.0 in /usr/local/lib/python3.10/dist-packages (from tensorflow==2.8.0) (1.6.3)\n",
            "Requirement already satisfied: flatbuffers>=1.12 in /usr/local/lib/python3.10/dist-packages (from tensorflow==2.8.0) (23.3.3)\n",
            "Requirement already satisfied: gast>=0.2.1 in /usr/local/lib/python3.10/dist-packages (from tensorflow==2.8.0) (0.4.0)\n",
            "Requirement already satisfied: google-pasta>=0.1.1 in /usr/local/lib/python3.10/dist-packages (from tensorflow==2.8.0) (0.2.0)\n",
            "Requirement already satisfied: h5py>=2.9.0 in /usr/local/lib/python3.10/dist-packages (from tensorflow==2.8.0) (3.8.0)\n",
            "Collecting keras-preprocessing>=1.1.1 (from tensorflow==2.8.0)\n",
            "  Downloading Keras_Preprocessing-1.1.2-py2.py3-none-any.whl (42 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m42.6/42.6 kB\u001b[0m \u001b[31m5.5 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hRequirement already satisfied: libclang>=9.0.1 in /usr/local/lib/python3.10/dist-packages (from tensorflow==2.8.0) (16.0.0)\n",
            "Requirement already satisfied: numpy>=1.20 in /usr/local/lib/python3.10/dist-packages (from tensorflow==2.8.0) (1.22.4)\n",
            "Requirement already satisfied: opt-einsum>=2.3.2 in /usr/local/lib/python3.10/dist-packages (from tensorflow==2.8.0) (3.3.0)\n",
            "Requirement already satisfied: protobuf>=3.9.2 in /usr/local/lib/python3.10/dist-packages (from tensorflow==2.8.0) (3.20.3)\n",
            "Requirement already satisfied: setuptools in /usr/local/lib/python3.10/dist-packages (from tensorflow==2.8.0) (67.7.2)\n",
            "Requirement already satisfied: six>=1.12.0 in /usr/local/lib/python3.10/dist-packages (from tensorflow==2.8.0) (1.16.0)\n",
            "Requirement already satisfied: termcolor>=1.1.0 in /usr/local/lib/python3.10/dist-packages (from tensorflow==2.8.0) (2.3.0)\n",
            "Requirement already satisfied: typing-extensions>=3.6.6 in /usr/local/lib/python3.10/dist-packages (from tensorflow==2.8.0) (4.5.0)\n",
            "Requirement already satisfied: wrapt>=1.11.0 in /usr/local/lib/python3.10/dist-packages (from tensorflow==2.8.0) (1.14.1)\n",
            "Collecting tensorboard<2.9,>=2.8 (from tensorflow==2.8.0)\n",
            "  Downloading tensorboard-2.8.0-py3-none-any.whl (5.8 MB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m5.8/5.8 MB\u001b[0m \u001b[31m37.7 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hCollecting tf-estimator-nightly==2.8.0.dev2021122109 (from tensorflow==2.8.0)\n",
            "  Downloading tf_estimator_nightly-2.8.0.dev2021122109-py2.py3-none-any.whl (462 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m462.5/462.5 kB\u001b[0m \u001b[31m38.2 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hCollecting keras<2.9,>=2.8.0rc0 (from tensorflow==2.8.0)\n",
            "  Downloading keras-2.8.0-py2.py3-none-any.whl (1.4 MB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m1.4/1.4 MB\u001b[0m \u001b[31m66.3 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hRequirement already satisfied: tensorflow-io-gcs-filesystem>=0.23.1 in /usr/local/lib/python3.10/dist-packages (from tensorflow==2.8.0) (0.32.0)\n",
            "Requirement already satisfied: grpcio<2.0,>=1.24.3 in /usr/local/lib/python3.10/dist-packages (from tensorflow==2.8.0) (1.54.0)\n",
            "Collecting coloredlogs (from onnxruntime==1.15.0)\n",
            "  Downloading coloredlogs-15.0.1-py2.py3-none-any.whl (46 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m46.0/46.0 kB\u001b[0m \u001b[31m5.9 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hRequirement already satisfied: packaging in /usr/local/lib/python3.10/dist-packages (from onnxruntime==1.15.0) (23.1)\n",
            "Requirement already satisfied: sympy in /usr/local/lib/python3.10/dist-packages (from onnxruntime==1.15.0) (1.11.1)\n",
            "Requirement already satisfied: wheel<1.0,>=0.23.0 in /usr/local/lib/python3.10/dist-packages (from astunparse>=1.6.0->tensorflow==2.8.0) (0.40.0)\n",
            "Requirement already satisfied: google-auth<3,>=1.6.3 in /usr/local/lib/python3.10/dist-packages (from tensorboard<2.9,>=2.8->tensorflow==2.8.0) (2.17.3)\n",
            "Collecting google-auth-oauthlib<0.5,>=0.4.1 (from tensorboard<2.9,>=2.8->tensorflow==2.8.0)\n",
            "  Downloading google_auth_oauthlib-0.4.6-py2.py3-none-any.whl (18 kB)\n",
            "Requirement already satisfied: markdown>=2.6.8 in /usr/local/lib/python3.10/dist-packages (from tensorboard<2.9,>=2.8->tensorflow==2.8.0) (3.4.3)\n",
            "Requirement already satisfied: requests<3,>=2.21.0 in /usr/local/lib/python3.10/dist-packages (from tensorboard<2.9,>=2.8->tensorflow==2.8.0) (2.27.1)\n",
            "Collecting tensorboard-data-server<0.7.0,>=0.6.0 (from tensorboard<2.9,>=2.8->tensorflow==2.8.0)\n",
            "  Downloading tensorboard_data_server-0.6.1-py3-none-manylinux2010_x86_64.whl (4.9 MB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m4.9/4.9 MB\u001b[0m \u001b[31m57.2 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hRequirement already satisfied: tensorboard-plugin-wit>=1.6.0 in /usr/local/lib/python3.10/dist-packages (from tensorboard<2.9,>=2.8->tensorflow==2.8.0) (1.8.1)\n",
            "Requirement already satisfied: werkzeug>=0.11.15 in /usr/local/lib/python3.10/dist-packages (from tensorboard<2.9,>=2.8->tensorflow==2.8.0) (2.3.0)\n",
            "Collecting humanfriendly>=9.1 (from coloredlogs->onnxruntime==1.15.0)\n",
            "  Downloading humanfriendly-10.0-py2.py3-none-any.whl (86 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m86.8/86.8 kB\u001b[0m \u001b[31m12.2 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hRequirement already satisfied: mpmath>=0.19 in /usr/local/lib/python3.10/dist-packages (from sympy->onnxruntime==1.15.0) (1.3.0)\n",
            "Requirement already satisfied: cachetools<6.0,>=2.0.0 in /usr/local/lib/python3.10/dist-packages (from google-auth<3,>=1.6.3->tensorboard<2.9,>=2.8->tensorflow==2.8.0) (5.3.0)\n",
            "Requirement already satisfied: pyasn1-modules>=0.2.1 in /usr/local/lib/python3.10/dist-packages (from google-auth<3,>=1.6.3->tensorboard<2.9,>=2.8->tensorflow==2.8.0) (0.3.0)\n",
            "Requirement already satisfied: rsa<5,>=3.1.4 in /usr/local/lib/python3.10/dist-packages (from google-auth<3,>=1.6.3->tensorboard<2.9,>=2.8->tensorflow==2.8.0) (4.9)\n",
            "Requirement already satisfied: requests-oauthlib>=0.7.0 in /usr/local/lib/python3.10/dist-packages (from google-auth-oauthlib<0.5,>=0.4.1->tensorboard<2.9,>=2.8->tensorflow==2.8.0) (1.3.1)\n",
            "Requirement already satisfied: urllib3<1.27,>=1.21.1 in /usr/local/lib/python3.10/dist-packages (from requests<3,>=2.21.0->tensorboard<2.9,>=2.8->tensorflow==2.8.0) (1.26.15)\n",
            "Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.10/dist-packages (from requests<3,>=2.21.0->tensorboard<2.9,>=2.8->tensorflow==2.8.0) (2022.12.7)\n",
            "Requirement already satisfied: charset-normalizer~=2.0.0 in /usr/local/lib/python3.10/dist-packages (from requests<3,>=2.21.0->tensorboard<2.9,>=2.8->tensorflow==2.8.0) (2.0.12)\n",
            "Requirement already satisfied: idna<4,>=2.5 in /usr/local/lib/python3.10/dist-packages (from requests<3,>=2.21.0->tensorboard<2.9,>=2.8->tensorflow==2.8.0) (3.4)\n",
            "Requirement already satisfied: MarkupSafe>=2.1.1 in /usr/local/lib/python3.10/dist-packages (from werkzeug>=0.11.15->tensorboard<2.9,>=2.8->tensorflow==2.8.0) (2.1.2)\n",
            "Requirement already satisfied: pyasn1<0.6.0,>=0.4.6 in /usr/local/lib/python3.10/dist-packages (from pyasn1-modules>=0.2.1->google-auth<3,>=1.6.3->tensorboard<2.9,>=2.8->tensorflow==2.8.0) (0.5.0)\n",
            "Requirement already satisfied: oauthlib>=3.0.0 in /usr/local/lib/python3.10/dist-packages (from requests-oauthlib>=0.7.0->google-auth-oauthlib<0.5,>=0.4.1->tensorboard<2.9,>=2.8->tensorflow==2.8.0) (3.2.2)\n",
            "Installing collected packages: tf-estimator-nightly, keras, tensorflow-hub, tensorboard-data-server, onnx, keras-preprocessing, humanfriendly, coloredlogs, onnxruntime, google-auth-oauthlib, tensorboard, tensorflow\n",
            "  Attempting uninstall: keras\n",
            "    Found existing installation: keras 2.12.0\n",
            "    Uninstalling keras-2.12.0:\n",
            "      Successfully uninstalled keras-2.12.0\n",
            "  Attempting uninstall: tensorflow-hub\n",
            "    Found existing installation: tensorflow-hub 0.13.0\n",
            "    Uninstalling tensorflow-hub-0.13.0:\n",
            "      Successfully uninstalled tensorflow-hub-0.13.0\n",
            "  Attempting uninstall: tensorboard-data-server\n",
            "    Found existing installation: tensorboard-data-server 0.7.0\n",
            "    Uninstalling tensorboard-data-server-0.7.0:\n",
            "      Successfully uninstalled tensorboard-data-server-0.7.0\n",
            "  Attempting uninstall: google-auth-oauthlib\n",
            "    Found existing installation: google-auth-oauthlib 1.0.0\n",
            "    Uninstalling google-auth-oauthlib-1.0.0:\n",
            "      Successfully uninstalled google-auth-oauthlib-1.0.0\n",
            "  Attempting uninstall: tensorboard\n",
            "    Found existing installation: tensorboard 2.12.2\n",
            "    Uninstalling tensorboard-2.12.2:\n",
            "      Successfully uninstalled tensorboard-2.12.2\n",
            "  Attempting uninstall: tensorflow\n",
            "    Found existing installation: tensorflow 2.12.0\n",
            "    Uninstalling tensorflow-2.12.0:\n",
            "      Successfully uninstalled tensorflow-2.12.0\n",
            "Successfully installed coloredlogs-15.0.1 google-auth-oauthlib-0.4.6 humanfriendly-10.0 keras-2.8.0 keras-preprocessing-1.1.2 onnx-1.14.0 onnxruntime-1.15.0 tensorboard-2.8.0 tensorboard-data-server-0.6.1 tensorflow-2.8.0 tensorflow-hub-0.12.0 tf-estimator-nightly-2.8.0.dev2021122109\n",
            "Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n",
            "Requirement already satisfied: gdown in /usr/local/lib/python3.10/dist-packages (4.6.6)\n",
            "Collecting gdown\n",
            "  Downloading gdown-4.7.1-py3-none-any.whl (15 kB)\n",
            "Requirement already satisfied: filelock in /usr/local/lib/python3.10/dist-packages (from gdown) (3.12.0)\n",
            "Requirement already satisfied: requests[socks] in /usr/local/lib/python3.10/dist-packages (from gdown) (2.27.1)\n",
            "Requirement already satisfied: six in /usr/local/lib/python3.10/dist-packages (from gdown) (1.16.0)\n",
            "Requirement already satisfied: tqdm in /usr/local/lib/python3.10/dist-packages (from gdown) (4.65.0)\n",
            "Requirement already satisfied: beautifulsoup4 in /usr/local/lib/python3.10/dist-packages (from gdown) (4.11.2)\n",
            "Requirement already satisfied: soupsieve>1.2 in /usr/local/lib/python3.10/dist-packages (from beautifulsoup4->gdown) (2.4.1)\n",
            "Requirement already satisfied: urllib3<1.27,>=1.21.1 in /usr/local/lib/python3.10/dist-packages (from requests[socks]->gdown) (1.26.15)\n",
            "Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.10/dist-packages (from requests[socks]->gdown) (2022.12.7)\n",
            "Requirement already satisfied: charset-normalizer~=2.0.0 in /usr/local/lib/python3.10/dist-packages (from requests[socks]->gdown) (2.0.12)\n",
            "Requirement already satisfied: idna<4,>=2.5 in /usr/local/lib/python3.10/dist-packages (from requests[socks]->gdown) (3.4)\n",
            "Requirement already satisfied: PySocks!=1.5.7,>=1.5.6 in /usr/local/lib/python3.10/dist-packages (from requests[socks]->gdown) (1.7.1)\n",
            "Installing collected packages: gdown\n",
            "  Attempting uninstall: gdown\n",
            "    Found existing installation: gdown 4.6.6\n",
            "    Uninstalling gdown-4.6.6:\n",
            "      Successfully uninstalled gdown-4.6.6\n",
            "Successfully installed gdown-4.7.1\n"
          ]
        }
      ],
      "source": [
        "!pip install tensorflow==2.8.0 tensorflow-hub==0.12.0 onnxruntime==1.15.0 onnx==1.14.0\n",
        "!pip install --upgrade gdown"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# DIS\n",
        "! git clone https://github.com/xuebinqin/DIS\n",
        "%cd /content/DIS/IS-Net\n",
        "\n",
        "!mkdir saved_models"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "AF9hHIttx-Nl",
        "outputId": "b65a2376-69a0-43be-97ad-0a1f65cc6d10"
      },
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Cloning into 'DIS'...\n",
            "remote: Enumerating objects: 328, done.\u001b[K\n",
            "remote: Counting objects: 100% (75/75), done.\u001b[K\n",
            "remote: Compressing objects: 100% (20/20), done.\u001b[K\n",
            "remote: Total 328 (delta 64), reused 56 (delta 55), pack-reused 253\u001b[K\n",
            "Receiving objects: 100% (328/328), 49.59 MiB | 23.02 MiB/s, done.\n",
            "Resolving deltas: 100% (153/153), done.\n",
            "/content/DIS/IS-Net\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "if remove_back=='ISNet':\n",
        "    import cv2\n",
        "    from data_loader_cache import normalize, im_reader, im_preprocess \n",
        "    import gdown\n",
        "    from google.colab.patches import cv2_imshow\n",
        "    from io import BytesIO\n",
        "    import matplotlib.pyplot as plt\n",
        "    from models import *\n",
        "    import numpy as np\n",
        "    import onnxruntime\n",
        "    from PIL import Image\n",
        "    import requests\n",
        "    import torch\n",
        "    from torch.autograd import Variable\n",
        "    import torch.nn.functional as F\n",
        "    from torchvision import transforms\n",
        "\n",
        "    os.chdir('/content/DIS/IS-Net')\n",
        "    device = 'cpu'\n",
        "    gdown.download(\n",
        "        'https://drive.google.com/uc?id=1KyMpRjewZdyYfxHPYcd-ZbanIXtin0Sn', \n",
        "        './saved_models/isnet.pth', \n",
        "        quiet=False,\n",
        "    )\n",
        "    gdown.download(\n",
        "        'https://drive.google.com/uc?id=1nV57qKuy--d5u1yvkng9aXW1KS4sOpOi', \n",
        "        './saved_models/isnet-general-use.pth', \n",
        "        quiet=False,\n",
        "    )\n",
        "\n",
        "\n",
        "    class GOSNormalize(object):\n",
        "        '''\n",
        "        Normalize the Image using torch.transforms\n",
        "        '''\n",
        "        def __init__(self, mean=[0.485,0.456,0.406], std=[0.229,0.224,0.225]):\n",
        "            self.mean = mean\n",
        "            self.std = std\n",
        "\n",
        "        def __call__(self,image):\n",
        "            image = normalize(image,self.mean,self.std)\n",
        "            return image\n",
        "    \n",
        "    def build_model(hypar,device):\n",
        "        net = hypar[\"model\"]#GOSNETINC(3,1)\n",
        "\n",
        "        # convert to half precision\n",
        "        if(hypar[\"model_digit\"]==\"half\"):\n",
        "            net.half()\n",
        "            for layer in net.modules():\n",
        "                if isinstance(layer, nn.BatchNorm2d):\n",
        "                    layer.float()\n",
        "\n",
        "        net.to(device)\n",
        "\n",
        "        if(hypar[\"restore_model\"]!=\"\"):\n",
        "            net.load_state_dict(torch.load(hypar[\"model_path\"]+\"/\"+hypar[\"restore_model\"],map_location=device))\n",
        "            net.to(device)\n",
        "        net.eval()  \n",
        "        return net\n",
        "\n",
        "    transform =  transforms.Compose([GOSNormalize([0.5,0.5,0.5],[1.0,1.0,1.0])])\n",
        "\n",
        "    # gen param\n",
        "    hypar = {} # paramters for inferencing\n",
        "\n",
        "    hypar[\"model_path\"] =\"./saved_models\" ## load trained weights from this path\n",
        "    hypar[\"restore_model\"] = \"isnet-general-use.pth\"  # \"isnet.pth\" ## name of the to-be-loaded weights\n",
        "    hypar[\"interm_sup\"] = False ## indicate if activate intermediate feature supervision\n",
        "\n",
        "    ##  choose floating point accuracy --\n",
        "    hypar[\"model_digit\"] = \"full\" ## indicates \"half\" or \"full\" accuracy of float number\n",
        "    hypar[\"seed\"] = 0\n",
        "\n",
        "    hypar[\"cache_size\"] = [1024, 1024] ## cached input spatial resolution, can be configured into different size\n",
        "\n",
        "    ## data augmentation parameters ---\n",
        "    hypar[\"input_size\"] = [1024, 1024] ## mdoel input spatial size, usually use the same value hypar[\"cache_size\"], which means we don't further resize the images\n",
        "    hypar[\"crop_size\"] = [1024, 1024] ## random crop size from the input, it is usually set as smaller than hypar[\"cache_size\"], e.g., [920,920] for data augmentation\n",
        "\n",
        "    hypar[\"model\"] = ISNetDIS()\n",
        "\n",
        "    net = build_model(hypar, device)\n",
        "\n",
        "\n",
        "\n",
        "    def load_image(im_path, hypar):\n",
        "        if im_path.startswith(\"http\"):\n",
        "            im_path = BytesIO(requests.get(im_path).content)\n",
        "\n",
        "        im = im_reader(im_path)\n",
        "        im, im_shp = im_preprocess(im, hypar[\"cache_size\"])\n",
        "        im = torch.divide(im,255.0)\n",
        "        shape = torch.from_numpy(np.array(im_shp))\n",
        "        return transform(im).unsqueeze(0), shape.unsqueeze(0) # make a batch of image, shape\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "        \n",
        "    def predict(net,  inputs_val, shapes_val, hypar, device):\n",
        "        '''\n",
        "        Given an Image, predict the mask\n",
        "        '''\n",
        "        net.eval()\n",
        "\n",
        "        if(hypar[\"model_digit\"]==\"full\"):\n",
        "            inputs_val = inputs_val.type(torch.FloatTensor)\n",
        "        else:\n",
        "            inputs_val = inputs_val.type(torch.HalfTensor)\n",
        "\n",
        "\n",
        "        inputs_val_v = Variable(inputs_val, requires_grad=False).to(device) # wrap inputs in Variable\n",
        "\n",
        "        ds_val = net(inputs_val_v)[0] # list of 6 results\n",
        "\n",
        "        pred_val = ds_val[0][0,:,:,:] # B x 1 x H x W    # we want the first one which is the most accurate prediction\n",
        "\n",
        "        ## recover the prediction spatial size to the orignal image size\n",
        "        pred_val = torch.squeeze(F.upsample(torch.unsqueeze(pred_val,0),(shapes_val[0][0],shapes_val[0][1]),mode='bilinear'))\n",
        "\n",
        "        ma = torch.max(pred_val)\n",
        "        mi = torch.min(pred_val)\n",
        "        pred_val = (pred_val-mi)/(ma-mi) # max = 1\n",
        "\n",
        "        if device == 'cuda': torch.cuda.empty_cache()\n",
        "        return (pred_val.detach().cpu().numpy()*255).astype(np.uint8) # it is the mask we need\n",
        "\n",
        "\n",
        "    # ONNX convert\n",
        "    def convert_to_onnx(net, file_name='output.onnx', input_shape=(512, 512), device='cpu'):\n",
        "        input = torch.randn(1, 3, input_shape[0], input_shape[1]).to(device)\n",
        "\n",
        "        input_layer_names = ['input_image']\n",
        "        output_layer_names = ['output_image']\n",
        "\n",
        "        torch.onnx.export(\n",
        "            net, \n",
        "            input, \n",
        "            file_name, \n",
        "            verbose=True,\n",
        "            opset_version=13,\n",
        "            input_names=input_layer_names,\n",
        "            output_names=output_layer_names,\n",
        "        )\n",
        "\n",
        "    convert_to_onnx(net, file_name='isnet-general-use.onnx', input_shape=(1024, 1024))\n",
        "\n",
        "\n",
        "    # inference\n",
        "    def run_inference(onnx_session, image, score_th=None):\n",
        "        # ONNX Input Size\n",
        "        input_size = onnx_session.get_inputs()[0].shape\n",
        "        input_width = input_size[3]\n",
        "        input_height = input_size[2]\n",
        "\n",
        "        # Pre process:Resize, BGR->RGB, float32 cast\n",
        "        input_image = cv2.resize(image, dsize=(input_width, input_height))\n",
        "        input_image = cv2.cvtColor(input_image, cv2.COLOR_BGR2RGB)\n",
        "        mean = [0.5, 0.5, 0.5]\n",
        "        std = [1.0, 1.0, 1.0]\n",
        "        input_image = (input_image / 255.0 - mean) / std\n",
        "        input_image = input_image.transpose(2, 0, 1)\n",
        "        input_image = np.expand_dims(input_image, axis=0)\n",
        "        input_image = input_image.astype('float32')\n",
        "\n",
        "        # Inference\n",
        "        input_name = onnx_session.get_inputs()[0].name\n",
        "        output_name = onnx_session.get_outputs()[0].name\n",
        "        result = onnx_session.run([output_name], {input_name: input_image})\n",
        "\n",
        "        # Post process:squeeze, Sigmoid, Normarize, uint8 cast\n",
        "        mask = np.squeeze(result[0])\n",
        "        min_value = np.min(mask)\n",
        "        max_value = np.max(mask)\n",
        "        mask = (mask - min_value) / (max_value - min_value)\n",
        "        if score_th is not None:\n",
        "            mask = np.where(mask < score_th, 0, 1)\n",
        "        mask *= 255\n",
        "        mask = mask.astype('uint8')\n",
        "\n",
        "        return mask"
      ],
      "metadata": {
        "id": "lKXKHbY8GxXG",
        "outputId": "fa80e151-cb6f-4ae6-d1cf-4e899106c399",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/torch/nn/_reduction.py:42: UserWarning: size_average and reduce args will be deprecated, please use reduction='mean' instead.\n",
            "  warnings.warn(warning.format(ret))\n",
            "Downloading...\n",
            "From (uriginal): https://drive.google.com/uc?id=1KyMpRjewZdyYfxHPYcd-ZbanIXtin0Sn\n",
            "From (redirected): https://drive.google.com/uc?id=1KyMpRjewZdyYfxHPYcd-ZbanIXtin0Sn&confirm=t&uuid=4a6276d9-44ab-4bf3-bb25-f4b558a3884d\n",
            "To: /content/DIS/IS-Net/saved_models/isnet.pth\n",
            "100%|██████████| 177M/177M [00:01<00:00, 140MB/s]\n",
            "Downloading...\n",
            "From (uriginal): https://drive.google.com/uc?id=1nV57qKuy--d5u1yvkng9aXW1KS4sOpOi\n",
            "From (redirected): https://drive.google.com/uc?id=1nV57qKuy--d5u1yvkng9aXW1KS4sOpOi&confirm=t&uuid=24188bc1-5638-473d-a70a-1b1a241a4799\n",
            "To: /content/DIS/IS-Net/saved_models/isnet-general-use.pth\n",
            "100%|██████████| 177M/177M [00:01<00:00, 160MB/s]\n",
            "/usr/local/lib/python3.10/dist-packages/torch/nn/functional.py:3737: UserWarning: nn.functional.upsample is deprecated. Use nn.functional.interpolate instead.\n",
            "  warnings.warn(\"nn.functional.upsample is deprecated. Use nn.functional.interpolate instead.\")\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "============= Diagnostic Run torch.onnx.export version 2.0.1+cu118 =============\n",
            "verbose: False, log level: Level.ERROR\n",
            "======================= 0 NONE 0 NOTE 0 WARNING 0 ERROR ========================\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "if remove_back=='ISNet':\n",
        "    os.chdir('/content/DIS/IS-Net')\n",
        "    image = cv2.imread('/content/upload.jpg')\n",
        "\n",
        "    onnx_session = onnxruntime.InferenceSession(\n",
        "        'isnet-general-use.onnx',\n",
        "        providers=[\n",
        "            'CUDAExecutionProvider',\n",
        "            'CPUExecutionProvider',\n",
        "        ],\n",
        "    )\n",
        "    mask = run_inference(\n",
        "        onnx_session,\n",
        "        image,\n",
        "        score_th=0.5,\n",
        "    )\n",
        "\n",
        "    mask = cv2.resize(\n",
        "        mask,\n",
        "        dsize=(image.shape[1], image.shape[0]),\n",
        "    )\n",
        "\n",
        "    overlay_image = np.zeros(image.shape, dtype=np.uint8)\n",
        "    overlay_image[:] = (255, 255, 255)\n",
        "    data = np.stack((mask, ) * 3, axis=-1).astype('uint8')\n",
        "    mask_image = np.where(data, image, overlay_image)\n",
        "\n",
        "    # 表示\n",
        "    cv2_imshow(data)\n",
        "    data = cv2.cvtColor(data, cv2.COLOR_BGR2GRAY)"
      ],
      "metadata": {
        "id": "WIr6RulmGszL",
        "outputId": "49388952-010b-460a-b59b-40183dd14f34",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 792
        }
      },
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/onnxruntime/capi/onnxruntime_inference_collection.py:65: UserWarning: Specified provider 'CUDAExecutionProvider' is not in available provider names.Available providers: 'CPUExecutionProvider'\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<PIL.Image.Image image mode=RGB size=550x720 at 0x7FED08301150>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 12,
      "metadata": {
        "id": "bxBmGrkbCNoT"
      },
      "outputs": [],
      "source": [
        "if remove_back=='Midas':\n",
        "    import cv2\n",
        "    import numpy as np\n",
        "\n",
        "    import matplotlib.pyplot as plt\n",
        "    import tensorflow_hub as hub\n",
        "    import tensorflow as tf\n",
        "\n",
        "    module = hub.load(\"https://tfhub.dev/intel/midas/v2_1_small/1\", tags=['serve'])\n",
        "    model = module.signatures['serving_default']\n",
        "\n",
        "\n",
        "    # 画像読み込み\n",
        "    image = cv2.imread('/content/upload.jpg')\n",
        "\n",
        "    # リサイズ\n",
        "    resize_image = cv2.resize(image, (256, 256))\n",
        "\n",
        "    # 正規化\n",
        "    resize_image = cv2.cvtColor(resize_image, cv2.COLOR_BGR2RGB) / 255.0\n",
        "\n",
        "    # 形状変更\n",
        "    resize_image = resize_image.transpose(2, 0, 1)\n",
        "    resize_image = resize_image.reshape(1, 3, 256, 256)\n",
        "\n",
        "    # tensor形式へ変換\n",
        "    tensor = tf.convert_to_tensor(resize_image, dtype=tf.float32)\n",
        "\n",
        "    result = model(tensor)\n",
        "    predict_result = result['default'].numpy()\n",
        "    predict_result = np.squeeze(predict_result)\n",
        "\n",
        "    # オリジナル画像のサイズにリサイズ\n",
        "    predict_result = cv2.resize(predict_result, (image.shape[1], image.shape[0]))\n",
        "\n",
        "    # 最大値が255になるよう変換\n",
        "    depth_max = predict_result.max()\n",
        "    data = ((predict_result / depth_max) * 255).astype(np.uint8)\n",
        "\n",
        "    # カラーマップ画像へ変換\n",
        "    color_map = cv2.applyColorMap(data, cv2.COLORMAP_TURBO)\n",
        "\n",
        "    # 表示\n",
        "    plt.imshow(color_map)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "tpEmhlY4Er2N"
      },
      "source": [
        "線画化\n",
        "\n",
        "必ず実行"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 13,
      "metadata": {
        "id": "SZ2-AmlMe7Kd"
      },
      "outputs": [],
      "source": [
        "class image_utils:\n",
        "    def median_filter(self, src, ksize):\n",
        "        d = int((ksize-1)/2)\n",
        "        h, w = src.shape[0], src.shape[1]\n",
        "        dst = src.copy()\n",
        "\n",
        "        for y in range(d, h - d):\n",
        "            for x in range(d, w - d):\n",
        "                dst[y][x] = np.median(src[y-d:y+d+1, x-d:x+d+1])\n",
        "        return dst\n",
        "\n",
        "    def export_edges_anime(self, image_raw: np.ndarray, depth_data: np.ndarray, median_ksize: int = 3, morph_g_ksize: int = 3, depth_threshold: int = 150, canny_threshold: int = 150):\n",
        "        mask = depth_data > depth_threshold\n",
        "        # maskを二値化\n",
        "        mask = mask.astype(np.uint8) * 255\n",
        "        # imageとmaskの合成\n",
        "        result = cv2.bitwise_and(image_raw, image_raw, mask=mask)\n",
        "\n",
        "        return self.anime2line(result, median_ksize=median_ksize, morph_g_ksize=morph_g_ksize, threshold=canny_threshold)\n",
        "\n",
        "    def anime2line(self, image_raw: np.ndarray, median_ksize: int = 3, morph_g_ksize: int = 3, threshold: int = 150):\n",
        "        edges = cv2.cvtColor(image_raw, cv2.COLOR_BGR2GRAY)\n",
        "\n",
        "        # Canny輪郭検出\n",
        "        edges = cv2.Canny(edges, threshold, threshold)\n",
        "\n",
        "        # MorphologyEx\n",
        "        if morph_g_ksize == 1:\n",
        "            pass\n",
        "        else:\n",
        "            kernel_giant = np.ones((morph_g_ksize, morph_g_ksize), np.uint8)\n",
        "            edges = cv2.morphologyEx(edges, cv2.MORPH_GRADIENT, kernel_giant)\n",
        "\n",
        "        # 色の反転\n",
        "        edges = cv2.bitwise_not(edges)\n",
        "        edges = edges.astype(np.uint8) * 255 * 255\n",
        "        edges = cv2.cvtColor(edges, cv2.COLOR_GRAY2BGR)\n",
        "\n",
        "        # メディアンフィルタ\n",
        "        if median_ksize == 1:\n",
        "            pass\n",
        "        else:\n",
        "            edges = self.median_filter(edges, ksize=median_ksize)\n",
        "\n",
        "        ret, output = cv2.threshold(edges, 200, 255, cv2.THRESH_BINARY)\n",
        "        return edges\n",
        "\n",
        "    # 白を透明にして色付きを半透明の白にする\n",
        "    def white2transparent(self, input_image: np.ndarray, os_type: str = 'Ubuntu'):\n",
        "        if os_type == 'Ubuntu':\n",
        "            image_png = cv2.cvtColor(input_image, cv2.COLOR_RGB2RGBA)\n",
        "            for i in range(image_png.shape[0]):\n",
        "                for j in range(image_png.shape[1]):\n",
        "                    if image_png[i][j][0] == 255 and image_png[i][j][1] == 255 and image_png[i][j][2] == 255:\n",
        "                        image_png[i][j][3] = 0\n",
        "                    else:\n",
        "                        image_png[i][j][0] = 255\n",
        "                        image_png[i][j][1] = 200\n",
        "                        image_png[i][j][2] = 255\n",
        "                        image_png[i][j][3] = 200\n",
        "        elif os_type == 'Windows':\n",
        "            image_png = cv2.cvtColor(input_image, cv2.COLOR_RGB2RGBA)\n",
        "            for i in range(image_png.shape[0]):\n",
        "                for j in range(image_png.shape[1]):\n",
        "                    if image_png[i][j][0] == 255 and image_png[i][j][1] == 255 and image_png[i][j][2] == 255:\n",
        "                        image_png[i][j][3] = 0\n",
        "                    else:\n",
        "                        image_png[i][j][0] = 200\n",
        "                        image_png[i][j][1] = 255\n",
        "                        image_png[i][j][2] = 255\n",
        "                        image_png[i][j][3] = 200\n",
        "        elif os_type == 'Mac-White':\n",
        "            image_png = cv2.cvtColor(input_image, cv2.COLOR_RGB2RGBA)\n",
        "            for i in range(image_png.shape[0]):\n",
        "                for j in range(image_png.shape[1]):\n",
        "                    if image_png[i][j][0] == 255 and image_png[i][j][1] == 255 and image_png[i][j][2] == 255:\n",
        "                        image_png[i][j][3] = 0\n",
        "                    else:\n",
        "                        image_png[i][j][0] = 200\n",
        "                        image_png[i][j][1] = 200\n",
        "                        image_png[i][j][2] = 200\n",
        "                        image_png[i][j][3] = 200\n",
        "        elif os_type == 'Mac-Gray':\n",
        "            image_png = cv2.cvtColor(input_image, cv2.COLOR_RGB2RGBA)\n",
        "            for i in range(image_png.shape[0]):\n",
        "                for j in range(image_png.shape[1]):\n",
        "                    if image_png[i][j][0] == 255 and image_png[i][j][1] == 255 and image_png[i][j][2] == 255:\n",
        "                        image_png[i][j][3] = 0\n",
        "                    else:\n",
        "                        image_png[i][j][0] = 70\n",
        "                        image_png[i][j][1] = 70\n",
        "                        image_png[i][j][2] = 70\n",
        "                        image_png[i][j][3] = 200\n",
        "        else:\n",
        "            print(\"cannot detect wallpaper OS\")\n",
        "        return image_png"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "5alWTXyxE4io"
      },
      "source": [
        "線画化２"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2ambD4SrfxEs"
      },
      "source": [
        "## Informative-drawings\n",
        "ファイルのダウンロード作業。スキップするほどでもないですが、ダウンロードを行わない場合はこの工程をスキップしてください。"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 14,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "iSLLrAFuE6aE",
        "outputId": "fbbde69e-491d-42a5-a347-ec60a93ce9b3"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Cloning into 'informative-drawings'...\n",
            "remote: Enumerating objects: 101, done.\u001b[K\n",
            "remote: Counting objects: 100% (28/28), done.\u001b[K\n",
            "remote: Compressing objects: 100% (17/17), done.\u001b[K\n",
            "remote: Total 101 (delta 21), reused 12 (delta 11), pack-reused 73\u001b[K\n",
            "Receiving objects: 100% (101/101), 3.86 MiB | 23.64 MiB/s, done.\n",
            "Resolving deltas: 100% (39/39), done.\n",
            "/content/DIS/IS-Net/informative-drawings\n",
            "Downloading...\n",
            "From (uriginal): https://drive.google.com/uc?id=11l5u5sb1PO5Z5YA3IoEHauVPm0k407C1\n",
            "From (redirected): https://drive.google.com/uc?id=11l5u5sb1PO5Z5YA3IoEHauVPm0k407C1&confirm=t&uuid=bee8b26b-d304-40db-bd8d-a50303b29175\n",
            "To: /content/DIS/IS-Net/informative-drawings/model.zip\n",
            "100% 47.7M/47.7M [00:00<00:00, 83.4MB/s]\n",
            "Archive:  model.zip\n",
            "   creating: checkpoints/model/\n",
            "   creating: checkpoints/model/contour_style/\n",
            "  inflating: checkpoints/model/contour_style/netG_A_latest.pth  \n",
            "   creating: checkpoints/model/opensketch_style/\n",
            "  inflating: checkpoints/model/opensketch_style/.DS_Store  \n",
            "  inflating: checkpoints/model/opensketch_style/netG_A_latest.pth  \n",
            "   creating: checkpoints/model/anime_style/\n",
            "  inflating: checkpoints/model/anime_style/netG_A_latest.pth  \n"
          ]
        }
      ],
      "source": [
        "!git clone https://github.com/carolineec/informative-drawings\n",
        "%cd informative-drawings\n",
        "\n",
        "!mkdir checkpoints\n",
        "!gdown 'https://drive.google.com/uc?id=11l5u5sb1PO5Z5YA3IoEHauVPm0k407C1'\n",
        "!unzip model.zip -d checkpoints"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 15,
      "metadata": {
        "id": "L5wMeGH-FPTV"
      },
      "outputs": [],
      "source": [
        "import torch\n",
        "from model import Generator\n",
        "\n",
        "import onnxruntime\n",
        "import cv2 as cv\n",
        "import numpy as np\n",
        "\n",
        "input_size = 512\n",
        "# ドットアクセス用の辞書クラス\n",
        "class DictDotNotation(dict): \n",
        "    def __init__(self, *args, **kwargs): \n",
        "        super().__init__(*args, **kwargs) \n",
        "        self.__dict__ = self\n",
        "# 疑似的なコマンドライン引数生成\n",
        "\n",
        "class InformativeDrawings:\n",
        "    def create_options(self) -> DictDotNotation:\n",
        "        opt = DictDotNotation()\n",
        "\n",
        "        opt.name = 'model/anime_style'\n",
        "        opt.dataroot = 'examples/test'\n",
        "        opt.checkpoints_dir = 'checkpoints'\n",
        "        opt.results_dir = 'results'\n",
        "        opt.geom_name = 'feats2Geom'\n",
        "        opt.batchSize = 1\n",
        "        opt.depthroot = ''\n",
        "        opt.input_nc = 3\n",
        "        opt.output_nc = 1\n",
        "        opt.geom_nc = 3\n",
        "        opt.every_feat = 1\n",
        "        opt.num_classes = 55\n",
        "        opt.midas = 0\n",
        "        opt.ngf = 64\n",
        "        opt.n_blocks = 3\n",
        "        opt.size = input_size\n",
        "        opt.cuda = True\n",
        "        opt.n_cpu = 8\n",
        "        opt.which_epoch = 'latest'\n",
        "        opt.aspect_ratio = 1.0\n",
        "        opt.mode = 'test'\n",
        "        opt.load_size = input_size\n",
        "        opt.crop_size = input_size\n",
        "        opt.max_dataset_size = float(\"inf\")\n",
        "        opt.preprocess = 'resize_and_crop'\n",
        "        opt.no_flip = False\n",
        "        opt.norm = 'instance'\n",
        "        opt.predict_depth = 0\n",
        "        opt.save_input = 0\n",
        "        opt.reconstruct = 0\n",
        "        opt.how_many = 100\n",
        "        return opt\n",
        "\n",
        "    def convert_to_onnx(self, net, file_name='output.onnx', input_shape=(512, 512), device='cpu'):\n",
        "        input = torch.randn(1, 3, input_shape[0], input_shape[1]).to(device)\n",
        "\n",
        "        input_layer_names = ['input_image']\n",
        "        output_layer_names = ['output_image']\n",
        "\n",
        "        torch.onnx.export(\n",
        "            net, \n",
        "            input, \n",
        "            file_name, \n",
        "            verbose=True,\n",
        "            opset_version=11,\n",
        "            input_names=input_layer_names,\n",
        "            output_names=output_layer_names,\n",
        "        )\n",
        "    \n",
        "    def export_edges_anime(self, image_raw: np.ndarray, depth_data: np.ndarray, morph_g_ksize: int = 3, depth_threshold: int = 150) -> np.ndarray:\n",
        "        mask = depth_data > depth_threshold\n",
        "        # maskを二値化\n",
        "        mask = mask.astype(np.uint8) * 255\n",
        "        # imageとmaskの合成\n",
        "        result = cv2.bitwise_and(image_raw, image_raw, mask=mask)\n",
        "\n",
        "        return self.generate(result, morph_g_ksize)\n",
        "\n",
        "    def generate(self, input_image, morph_g_ksize) -> np.ndarray:\n",
        "\n",
        "        # ネットワーク生成\n",
        "        opt = self.create_options()\n",
        "        net_G = Generator(opt.input_nc, opt.output_nc, opt.n_blocks)\n",
        "\n",
        "        # 重みをロード\n",
        "        net_G.load_state_dict(torch.load(os.path.join(opt.checkpoints_dir, opt.name, 'netG_A_%s.pth' % opt.which_epoch)))\n",
        "        device = 'cpu'\n",
        "        net_G.eval().to(device)\n",
        "\n",
        "        print('Loaded：', os.path.join(opt.checkpoints_dir, opt.name, 'netG_A_%s.pth' % opt.which_epoch))\n",
        "\n",
        "        self.convert_to_onnx(net_G, 'anime_style_512x512.onnx', (input_size, input_size), device)\n",
        "\n",
        "        # ONNXモデルロード\n",
        "        onnx_session = onnxruntime.InferenceSession(\n",
        "            'anime_style_512x512.onnx',\n",
        "            providers=['CUDAExecutionProvider', 'CPUExecutionProvider'],\n",
        "        )\n",
        "\n",
        "        input_name = onnx_session.get_inputs()[0].name\n",
        "        output_name = onnx_session.get_outputs()[0].name\n",
        "\n",
        "\n",
        "        # 前処理 (input is image)\n",
        "        image_data = cv.resize(input_image, dsize=(input_size, input_size))\n",
        "        x = cv.cvtColor(image_data, cv.COLOR_BGR2RGB)\n",
        "\n",
        "        x = np.array(x, dtype=np.float32)\n",
        "        x = x.transpose(2, 0, 1)\n",
        "        x = x.reshape(-1, 3, input_size, input_size)\n",
        "\n",
        "        # 推論\n",
        "        onnx_result = onnx_session.run([output_name], {input_name: x})\n",
        "\n",
        "        # 後処理\n",
        "        onnx_result = np.array(onnx_result).squeeze()\n",
        "        onnx_result = onnx_result * 255\n",
        "        onnx_result = onnx_result.astype(np.uint8)\n",
        "\n",
        "        \n",
        "        ret, output = cv2.threshold(onnx_result, 200, 255, cv2.THRESH_BINARY)\n",
        "\n",
        "        # kernel_giant = np.ones((morph_g_ksize, morph_g_ksize), np.uint8)\n",
        "        # output = cv2.morphologyEx(output, cv2.MORPH_GRADIENT, kernel_giant)\n",
        "\n",
        "        output = cv2.cvtColor(output, cv2.COLOR_GRAY2BGR)\n",
        "        output = cv2.resize(output, (image.shape[1],image.shape[0]))\n",
        "        return output"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 16,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "akicp1XSYg7C",
        "outputId": "b1ec209b-d2d2-4b8e-b23a-3ddb3283fe81"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "using informative drawings...\n",
            "Loaded： checkpoints/model/anime_style/netG_A_latest.pth\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/torch/onnx/_internal/jit_utils.py:306: UserWarning: Constant folding - Only steps=1 can be constant folded for opset >= 10 onnx::Slice op. Constant folding not applied. (Triggered internally at ../torch/csrc/jit/passes/onnx/constant_fold.cpp:179.)\n",
            "  _C._jit_pass_onnx_node_shape_type_inference(node, params_dict, opset_version)\n",
            "/usr/local/lib/python3.10/dist-packages/torch/onnx/symbolic_helper.py:1466: UserWarning: ONNX export mode is set to TrainingMode.EVAL, but operator 'instance_norm' is set to train=True. Exporting with train=True.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/onnx/utils.py:689: UserWarning: Constant folding - Only steps=1 can be constant folded for opset >= 10 onnx::Slice op. Constant folding not applied. (Triggered internally at ../torch/csrc/jit/passes/onnx/constant_fold.cpp:179.)\n",
            "  _C._jit_pass_onnx_graph_shape_type_inference(\n",
            "/usr/local/lib/python3.10/dist-packages/torch/onnx/utils.py:1186: UserWarning: Constant folding - Only steps=1 can be constant folded for opset >= 10 onnx::Slice op. Constant folding not applied. (Triggered internally at ../torch/csrc/jit/passes/onnx/constant_fold.cpp:179.)\n",
            "  _C._jit_pass_onnx_graph_shape_type_inference(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "============= Diagnostic Run torch.onnx.export version 2.0.1+cu118 =============\n",
            "verbose: False, log level: Level.ERROR\n",
            "======================= 0 NONE 0 NOTE 0 WARNING 0 ERROR ========================\n",
            "\n"
          ]
        }
      ],
      "source": [
        "image_utils_class = image_utils()\n",
        "\n",
        "if using_informative_drawings == \"True\":\n",
        "    print(\"using informative drawings...\")\n",
        "    id_class = InformativeDrawings()\n",
        "    os.chdir('/content/DIS/IS-Net/informative-drawings/')\n",
        "    image_line = id_class.export_edges_anime(image, depth_data=data, morph_g_ksize=morph_g_ksize, depth_threshold=depth_threshold)\n",
        "else:\n",
        "    print(\"using canny...\")\n",
        "    image_line = image_utils_class.export_edges_anime(image_raw=image, depth_data=data, median_ksize=median_ksize, morph_g_ksize=morph_g_ksize, depth_threshold=depth_threshold, canny_threshold=canny_threshold)\n",
        "#@markdown ## export_edges_anime関数\n",
        "#@markdown ### 引数一覧\n",
        "#@markdown - image_raw: 画像（読み込み時のまま）\n",
        "#@markdown - depth_data: 深度データ(2次元のndarray・0~255)\n",
        "#@markdown - threshold: 深度の閾値（0~255・int型）\n",
        "#@markdown\n",
        "#@markdown ### 出力一覧\n",
        "#@markdown - image: マスク済み画像\n",
        "#@markdown\n",
        "#@markdown ### 分岐\n",
        "#@markdown - informative drawings をTrueに設定するとAI線画に、FalseにするとCanny線画になります。"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 17,
      "metadata": {
        "id": "q-H83_5UJC54"
      },
      "outputs": [],
      "source": [
        "image_png = image_utils_class.white2transparent(image_line, OS)\n",
        "\n",
        "#@markdown ## white2transparent関数\n",
        "#@markdown 白領域を透明化する関数（色付きは半透明の白になります）"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 18,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 452
        },
        "id": "3pANOWzibWpL",
        "outputId": "15168671-2069-40e9-c8c9-804305294ba6"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.image.AxesImage at 0x7fed083021d0>"
            ]
          },
          "metadata": {},
          "execution_count": 18
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "plt.imshow(image_png)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 19,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 17
        },
        "id": "Kg-Zb97tdjBe",
        "outputId": "66b64b70-7ccb-440d-d69b-0508e1072dbf"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.Javascript object>"
            ],
            "application/javascript": [
              "\n",
              "    async function download(id, filename, size) {\n",
              "      if (!google.colab.kernel.accessAllowed) {\n",
              "        return;\n",
              "      }\n",
              "      const div = document.createElement('div');\n",
              "      const label = document.createElement('label');\n",
              "      label.textContent = `Downloading \"${filename}\": `;\n",
              "      div.appendChild(label);\n",
              "      const progress = document.createElement('progress');\n",
              "      progress.max = size;\n",
              "      div.appendChild(progress);\n",
              "      document.body.appendChild(div);\n",
              "\n",
              "      const buffers = [];\n",
              "      let downloaded = 0;\n",
              "\n",
              "      const channel = await google.colab.kernel.comms.open(id);\n",
              "      // Send a message to notify the kernel that we're ready.\n",
              "      channel.send({})\n",
              "\n",
              "      for await (const message of channel.messages) {\n",
              "        // Send a message to notify the kernel that we're ready.\n",
              "        channel.send({})\n",
              "        if (message.buffers) {\n",
              "          for (const buffer of message.buffers) {\n",
              "            buffers.push(buffer);\n",
              "            downloaded += buffer.byteLength;\n",
              "            progress.value = downloaded;\n",
              "          }\n",
              "        }\n",
              "      }\n",
              "      const blob = new Blob(buffers, {type: 'application/binary'});\n",
              "      const a = document.createElement('a');\n",
              "      a.href = window.URL.createObjectURL(blob);\n",
              "      a.download = filename;\n",
              "      div.appendChild(a);\n",
              "      a.click();\n",
              "      div.remove();\n",
              "    }\n",
              "  "
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.Javascript object>"
            ],
            "application/javascript": [
              "download(\"download_7cb1b057-5f88-42a2-b2b8-362ed605df7a\", \"result.png\", 19931)"
            ]
          },
          "metadata": {}
        }
      ],
      "source": [
        "plt.imsave(\"/content/result.png\",image_png)\n",
        "from google.colab import files\n",
        "files.download('/content/result.png')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 20,
      "metadata": {
        "id": "nU4eubO3EveY"
      },
      "outputs": [],
      "source": [
        "# Export depth\n",
        "#@markdown convert.pyで深度情報を確認する時はこのセルを実行して`depth_result.png`をダウンロードします。\n",
        "\n",
        "# # v----この行以下をコメントアウト--------v\n",
        "# export_png = np.zeros((data.shape[0], data.shape[1], 3))\n",
        "# export_png[:,:,0] = data/255\n",
        "# export_png[:,:,1] = data/255\n",
        "# export_png[:,:,2] = data/255\n",
        "\n",
        "# plt.imsave(\"/content/depth_result.png\",export_png)\n",
        "# files.download('/content/depth_result.png')"
      ]
    }
  ],
  "metadata": {
    "accelerator": "GPU",
    "colab": {
      "provenance": []
    },
    "kernelspec": {
      "display_name": "Python 3.10.6 64-bit ('3.10.6')",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "name": "python",
      "version": "3.10.6"
    },
    "vscode": {
      "interpreter": {
        "hash": "b4befac3e56e39e73ea575fd36aeeafdc8ef41834eee22ae63e330239ebc486c"
      }
    },
    "gpuClass": "standard"
  },
  "nbformat": 4,
  "nbformat_minor": 0
}