3dml_habr_phygitalism_part_1.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "w4xEsQU2C24A"
   },
   "source": [
    "## Установка необходимых для работы библиотек\n",
    "Необходимо выбрать `GPU` runtime type в настройках `runtime` google colab.\n",
    "Обратите внимание, что установка может занять несколько минут (обычно 5 минут)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 731
    },
    "colab_type": "code",
    "id": "cDRqZFmYRJ0V",
    "outputId": "eb4313f9-0427-4ac5-c19d-fbf81a8e510f"
   },
   "outputs": [],
   "source": [
    "!pip install mesh_to_sdf pyrender trimesh\n",
    "!pip install torch torchvision\n",
    "!pip install git+https://github.com/facebookresearch/pytorch3d.git@v0.2.0"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "OV_xLCxACsrd"
   },
   "source": [
    "## Загрузка библиотек"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "Dnu-swO7RMuA"
   },
   "outputs": [],
   "source": [
    "import os\n",
    "os.environ['PYOPENGL_PLATFORM'] = 'egl'\n",
    "\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from mpl_toolkits.mplot3d import Axes3D\n",
    "%matplotlib inline\n",
    "\n",
    "import torch\n",
    "\n",
    "# untilitis \n",
    "from pytorch3d.utils import ico_sphere\n",
    "\n",
    "# io utils\n",
    "from pytorch3d.io import load_obj\n",
    "\n",
    "# operations with data\n",
    "from pytorch3d.ops import sample_points_from_meshes\n",
    "\n",
    "# datastructures\n",
    "from pytorch3d.structures import Meshes, Textures\n",
    "\n",
    "# render \n",
    "from pytorch3d.renderer import (\n",
    "    look_at_view_transform,\n",
    "    OpenGLPerspectiveCameras,\n",
    "    DirectionalLights, \n",
    "    RasterizationSettings, \n",
    "    MeshRenderer, \n",
    "    MeshRasterizer,  \n",
    "    HardPhongShader\n",
    ")\n",
    "\n",
    "# mesh_to_sdf lib by marian42\n",
    "from mesh_to_sdf import mesh_to_sdf, sample_sdf_near_surface\n",
    "\n",
    "# trimesh lib imports\n",
    "import trimesh\n",
    "from trimesh.voxel.creation import voxelize"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "6cT6_RrhGCnV"
   },
   "outputs": [],
   "source": [
    "# If you have got a CUDA device, you can use GPU mode\n",
    "if torch.cuda.is_available():\n",
    "  device = torch.device('cuda:0')\n",
    "  torch.cuda.set_device(device)\n",
    "else:\n",
    "  device = torch.device('cpu')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "YE1Zv2JQZJ-M"
   },
   "source": [
    "# Полигональная модель"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "n4a2pg24ZYRK"
   },
   "source": [
    "Загрузка полигональной модели `bunny.obj` с сайта [Stanford graphics](https://graphics.stanford.edu/~mdfisher/Data/Meshes/bunny.obj), в случае, если работаем в colab."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 204
    },
    "colab_type": "code",
    "id": "QC7vBd_kjMHZ",
    "outputId": "9def6ee5-470e-4b06-e585-d49d35f5a474"
   },
   "outputs": [],
   "source": [
    "!mkdir data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "!wget -nc -P ./data/ https://graphics.stanford.edu/~mdfisher/Data/Meshes/bunny.obj"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "JEYkiCmwUe6p"
   },
   "outputs": [],
   "source": [
    "path_to_model = os.path.join(\"data\",\"bunny.obj\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "WHv3G8bpGQZI"
   },
   "source": [
    "### Создание полигональной икосферы sphere.obj с 4 уровнями subdivision"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "pnfg69AMGRjv"
   },
   "outputs": [],
   "source": [
    "# Trimesh sphere .obj model\n",
    "trimesh_sphere = trimesh.primitives.Sphere(subdivisions= 4)\n",
    "\n",
    "# Sphere mesh in pytorch3d\n",
    "sphere_mesh = ico_sphere(4, device)\n",
    "\n",
    "verts_rgb = torch.ones_like(sphere_mesh.verts_list()[0])[None]\n",
    "\n",
    "# Rainbow sphere in pytorch3d\n",
    "# verts_rgb = torch.rand_like(sphere_mesh.verts_list()[0])[None]\n",
    "\n",
    "sphere_mesh.textures = Textures(verts_rgb=verts_rgb.to(device))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "zufrySJrGcsN"
   },
   "source": [
    "### Создание полигональной модели кролика bunny.obj"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 51
    },
    "colab_type": "code",
    "id": "bwvu0Xs-GenD",
    "outputId": "f09ff737-b705-4a01-e436-61ce38ed14e1"
   },
   "outputs": [],
   "source": [
    "# Trimesh bunny .obj model\n",
    "bunny_trimesh = trimesh.load(path_to_model)\n",
    "\n",
    "if isinstance(bunny_trimesh, trimesh.Scene):\n",
    "    bunny_trimesh = bunny_trimesh.dump(concatenate=True)\n",
    "\n",
    "bunny_trimesh.vertices -= bunny_trimesh.center_mass\n",
    "scaling = 2 / bunny_trimesh.scale\n",
    "bunny_trimesh.apply_scale(scaling=scaling)\n",
    "\n",
    "# Rainbow bunny in trimesh\n",
    "# for facet in bunny_trimesh.facets:\n",
    "#   bunny_trimesh.visual.face_colors[facet] = trimesh.visual.random_color() \n",
    "\n",
    "# Bunny mesh in pytorch3d\n",
    "verts, faces_idx, _ = load_obj(path_to_model)\n",
    "faces = faces_idx.verts_idx\n",
    "\n",
    "center = verts.mean(0)\n",
    "verts = verts - center\n",
    "scale = max(verts.abs().max(0)[0])\n",
    "verts = verts / scale\n",
    "\n",
    "# Initialize each vertex to be white in color.\n",
    "verts_rgb = torch.ones_like(verts)[None]  # (1, V, 3)\n",
    "\n",
    "# Rainbow bunny in pytorch3d\n",
    "# verts_rgb = torch.rand_like(verts)[None]\n",
    "\n",
    "textures = Textures(verts_rgb=verts_rgb.to(device))\n",
    "\n",
    "# Create a Meshes object for the bunny.\n",
    "bunny_mesh = Meshes(\n",
    "    verts=[verts.to(device)],   \n",
    "    faces=[faces.to(device)], \n",
    "    textures=textures\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "xFucx_wPG6I7"
   },
   "source": [
    "### Преобразование полигональной модели к облаку точек с нормалями"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "noITD6T4HGOu"
   },
   "source": [
    "Один из основных способов создания модели облака точек из существующей полигональной модели - равномерное семплирование дополнительных вершин на гранях полигональной модели. \n",
    "\n",
    "Эту задачу можно решать как встроеными методами библиотек - `sample_points_from_meshes` в *pytorch3d*, или же можно сакмому написать алгоритм семплирования и использовать его."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "0vy1g0v5G7VX"
   },
   "outputs": [],
   "source": [
    "# Mesh to pointcloud with normals in pytorch3d\n",
    "num_points_to_sample = 25000\n",
    "\n",
    "bunny_vert, bunny_norm = sample_points_from_meshes(\n",
    "    bunny_mesh,\n",
    "    num_points_to_sample ,\n",
    "    return_normals=True\n",
    ")\n",
    "\n",
    "sphere_vert, sphere_norm = sample_points_from_meshes(\n",
    "    sphere_mesh,\n",
    "    num_points_to_sample,\n",
    "    return_normals=True\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "SqN3lQwEIIwS"
   },
   "source": [
    "# Блок визуализации"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "TO0hEPHrIJ-2"
   },
   "source": [
    "Визуализация облака точек, полученного из полигональной модели методом семплирования точек на гранях обоими методами."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "KN2O7kY3G-0L"
   },
   "outputs": [],
   "source": [
    "def plot_pointcloud(points, elev=70, azim=-70, title=\"\"):\n",
    "    # Sample points uniformly from the surface of the mesh.\n",
    "    fig = plt.figure(figsize=(10, 10))\n",
    "    ax = Axes3D(fig)\n",
    "    x, y, z = points\n",
    "    ax.scatter3D(x, z, -y,marker='.')  \n",
    "    ax.set_xlabel('x')\n",
    "    ax.set_ylabel('z')\n",
    "    ax.set_zlabel('y')\n",
    "    ax.set_title(title)\n",
    "    ax.view_init(elev, azim)\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "eR1stzzQSwZP"
   },
   "source": [
    "Визуализация моделей с помощью дифференциального рендера pytorch3d"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "B-leNp_cSx-e"
   },
   "outputs": [],
   "source": [
    "# Initialize an OpenGL perspective camera.\n",
    "cameras = OpenGLPerspectiveCameras(device=device)\n",
    "\n",
    "# We will also create a phong renderer. This is simpler and only needs to render one face per pixel.\n",
    "raster_settings = RasterizationSettings(\n",
    "    image_size=1024, \n",
    "    blur_radius=0, \n",
    "    faces_per_pixel=1, \n",
    ")\n",
    "\n",
    "# We can add a directional light in the scene. \n",
    "ambient_color = torch.FloatTensor([[0.0, 0.0, 0.0]]).to(device)\n",
    "diffuse_color = torch.FloatTensor([[1.0, 1.0, 1.0]]).to(device)\n",
    "specular_color = torch.FloatTensor([[0.1, 0.1, 0.1]]).to(device)\n",
    "direction = torch.FloatTensor([[1, 1, 1]]).to(device)\n",
    "lights = DirectionalLights(ambient_color=ambient_color,\n",
    "                           diffuse_color=diffuse_color,\n",
    "                           specular_color=specular_color,\n",
    "                           direction=direction,\n",
    "                           device=device)\n",
    "\n",
    "phong_renderer = MeshRenderer(\n",
    "    rasterizer=MeshRasterizer(\n",
    "        cameras=cameras, \n",
    "        raster_settings=raster_settings\n",
    "    ),\n",
    "    shader=HardPhongShader(\n",
    "        device=device, \n",
    "        cameras=cameras, \n",
    "        lights=lights\n",
    "        )\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "si6psBxCS1gF"
   },
   "outputs": [],
   "source": [
    "# Select the viewpoint using spherical angles  \n",
    "distance = 2.0   # distance from camera to the object`\n",
    "elevation = 40.0   # angle of elevation in degrees\n",
    "azimuth = 0.0  # No rotation so the camera is positioned on the +Z axis. \n",
    "\n",
    "# Get the position of the camera based on the spherical angles\n",
    "R, T = look_at_view_transform(distance, elevation, azimuth, device=device,at=((-0.02,0.1,0.0),))\n",
    "\n",
    "# Render the bunny providing the values of R and T. \n",
    "image_bunny = phong_renderer(meshes_world=bunny_mesh, R=R, T=T)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "e6Dphh9hS41L"
   },
   "outputs": [],
   "source": [
    "# Select the viewpoint using spherical angles  \n",
    "distance = 3.0   # distance from camera to the object`\n",
    "elevation = 40.0   # angle of elevation in degrees\n",
    "azimuth = 0.0  # No rotation so the camera is positioned on the +Z axis. \n",
    "\n",
    "# Get the position of the camera based on the spherical angles\n",
    "R, T = look_at_view_transform(distance, elevation, azimuth, device=device,at=((-0.02,0.1,0.0),))\n",
    "\n",
    "# Render the sphere providing the values of R and T. \n",
    "image_sphere = phong_renderer(meshes_world=sphere_mesh, R=R, T=T)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "CaRz2KdyS_yC"
   },
   "source": [
    "# Полигональные модели"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "0wszxJFCTAwW"
   },
   "source": [
    "### Визуализация в pytorch3d"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 775
    },
    "colab_type": "code",
    "id": "WUDsBuxfTFRq",
    "outputId": "96e384f9-a0fa-42b2-f51c-53805bb88156"
   },
   "outputs": [],
   "source": [
    "image_sphere = image_sphere.cpu().numpy()\n",
    "\n",
    "plt.figure(figsize=(13, 13))\n",
    "plt.imshow(image_sphere.squeeze())\n",
    "plt.grid(False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 775
    },
    "colab_type": "code",
    "id": "DhBA68GHTL54",
    "outputId": "f72ad1b5-0aa8-4fd7-89d4-d3662d7b8246"
   },
   "outputs": [],
   "source": [
    "image_bunny = image_bunny.cpu().numpy()\n",
    "\n",
    "plt.figure(figsize=(13, 13))\n",
    "plt.imshow(image_bunny.squeeze())\n",
    "plt.grid(False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "wBi69zGHTRoq"
   },
   "source": [
    "### Интерактивная визуализация в trimesh"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 521
    },
    "colab_type": "code",
    "id": "LJdhWgV2TS0d",
    "outputId": "10d906eb-3d6d-45a9-e607-d9242858aa37"
   },
   "outputs": [],
   "source": [
    "trimesh_sphere.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 521
    },
    "colab_type": "code",
    "id": "LpMRBHMSTWjR",
    "outputId": "f21b4b47-480a-4520-c4ce-8099ab7715da"
   },
   "outputs": [],
   "source": [
    "bunny_trimesh.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "WFTkoVEzTen8"
   },
   "source": [
    " ### Для полигональных моделей как для пространственных графов можно вычислить ряд геометрических характеристик."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 51
    },
    "colab_type": "code",
    "id": "_dTnmHdTTi_N",
    "outputId": "735d1dc6-ce36-4bbb-95c1-da3e85bc8aa7"
   },
   "outputs": [],
   "source": [
    "print(\n",
    "    \"Эйлерова характеристика пространственного графа модели bunny Xi = V - E + F =\",\n",
    "    bunny_trimesh.euler_number\n",
    ")\n",
    "\n",
    "print(\n",
    "    \"Эйлерова характеристика пространственного графа модели sphere Xi = V - E + F =\",\n",
    "    trimesh_sphere.euler_number\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 51
    },
    "colab_type": "code",
    "id": "CZsuIdlhTkW-",
    "outputId": "5fd34ed6-2eb2-4860-f75b-48ead189917a"
   },
   "outputs": [],
   "source": [
    "print(\"Is bunny mesh watertight:\", bunny_trimesh.is_watertight)\n",
    "print(\"Is sphere mesh watertight:\", trimesh_sphere.is_watertight)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 51
    },
    "colab_type": "code",
    "id": "iLuh1PCeTkmg",
    "outputId": "084c868b-9bf1-4fdb-d424-cc59b98dfd92"
   },
   "outputs": [],
   "source": [
    "print(\"Объем модели bunny.obj:\", bunny_trimesh.volume)\n",
    "print(\"Объем модели sphere.obj:\", trimesh_sphere.volume)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "id": "bZs2TJPiTpcA",
    "outputId": "826d7998-bc7a-4f87-e572-0cc05013a59c"
   },
   "outputs": [],
   "source": [
    "# Объем единичной сферы 4/3 * Pi\n",
    "(4/3)*np.pi"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "VHP3O_LkTvZs"
   },
   "source": [
    "# Воксельное представление"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "bSD8XFOqUUhm"
   },
   "outputs": [],
   "source": [
    "low_idx_bunny = bunny_trimesh.scale / 15\n",
    "high_idx_bunny = bunny_trimesh.scale / 39\n",
    "low_idx_sphere = trimesh_sphere.scale / 15\n",
    "high_idx_sphere = trimesh_sphere.scale / 30"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "-qHYWHYUUUsJ"
   },
   "outputs": [],
   "source": [
    "vox_high_bunny = voxelize(bunny_trimesh,pitch=high_idx_bunny)\n",
    "vox_high_sphere = voxelize(trimesh_sphere,pitch=high_idx_sphere)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 357
    },
    "colab_type": "code",
    "id": "uXZg-xCVOgB9",
    "outputId": "ca60c6f9-afca-4fde-ccad-3d3592ab96f5"
   },
   "outputs": [],
   "source": [
    "print(\"Срез воксельной сетки:\\n\",np.array(vox_high_sphere.matrix, dtype=np.uint8)[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "id": "mAh8Dyt-UUz4",
    "outputId": "4bb0e27e-8843-4639-877c-f2bf1958d743"
   },
   "outputs": [],
   "source": [
    "print(\"Размер воксельной сетки с высоким разрешением:\", vox_high_sphere.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 521
    },
    "colab_type": "code",
    "id": "D5vIZrDPUU8G",
    "outputId": "72456499-f2bd-4016-f248-ace3d6927a7c"
   },
   "outputs": [],
   "source": [
    "vox_high_sphere.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 521
    },
    "colab_type": "code",
    "id": "mizc1UN_Ulaa",
    "outputId": "3f0e909b-fad8-4f6a-bfed-603bc7dc0488"
   },
   "outputs": [],
   "source": [
    "vox_high_bunny.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "xdHKmFSPUlig"
   },
   "outputs": [],
   "source": [
    "vox_low_bunny = voxelize(bunny_trimesh,pitch=low_idx_bunny)\n",
    "vox_low_sphere = voxelize(trimesh_sphere,pitch=low_idx_sphere)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "id": "hddEieCRUlot",
    "outputId": "fd3702ce-2335-4db3-f223-423ff9ece59c"
   },
   "outputs": [],
   "source": [
    "print(\"Размер воксельной сетки с низким разрешением:\", vox_low_sphere.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 521
    },
    "colab_type": "code",
    "id": "4Gik0etTUtH1",
    "outputId": "a2ad2f42-5e2e-4949-b770-cff74a092343"
   },
   "outputs": [],
   "source": [
    "vox_low_sphere.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 521
    },
    "colab_type": "code",
    "id": "nyR5I0i-UtVp",
    "outputId": "61bca813-9499-4057-e41b-63b7e73d61f7"
   },
   "outputs": [],
   "source": [
    "vox_low_bunny.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "ux7udsuYU4mc"
   },
   "source": [
    "# Вычисление SDF функции"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "M0Ca67tCVEUu"
   },
   "source": [
    "Вычисление SDF функции для модели bunny.obj в двух точках: одной внутренней (центр масс), одной внешней."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "XLgguAlkVI18"
   },
   "outputs": [],
   "source": [
    "center_mass = bunny_trimesh.center_mass\n",
    "query_points = np.array([[center_mass],[[3,3,3]]])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 51
    },
    "colab_type": "code",
    "id": "knlEbGkcVLST",
    "outputId": "6c516819-a90d-4e59-a14e-ed1dc95c9a20"
   },
   "outputs": [],
   "source": [
    "for point in query_points:\n",
    "    print(\"SDF{0} = {1}\".format(point[0],mesh_to_sdf(bunny_trimesh,point)[0]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "e7CdrabwIYXQ"
   },
   "source": [
    "Вычисление SDF функции для модели sphere.obj в двух точках: одной внутренней (центр масс), одной внешней."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "829ynBXoVhl0"
   },
   "outputs": [],
   "source": [
    "center_mass = trimesh_sphere.center_mass\n",
    "query_points = np.array([[center_mass],[[3,3,3]]])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 51
    },
    "colab_type": "code",
    "id": "VO9BP2zYVhpP",
    "outputId": "0dda200c-6892-4cff-e9a4-a0473636d98b"
   },
   "outputs": [],
   "source": [
    "for point in query_points:\n",
    "    print(\"SDF{0} = {1}\".format(point[0],mesh_to_sdf(trimesh_sphere,point)[0]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "QKeuYtIlWHOH"
   },
   "source": [
    "Визуализируем "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "_JaIwowyRyP_"
   },
   "outputs": [],
   "source": [
    "points, sdf = sample_sdf_near_surface(trimesh_sphere, number_of_points=5000)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 1000
    },
    "colab_type": "code",
    "id": "GMewl4RMTz6m",
    "outputId": "00a1a27e-18d4-4a68-b213-5f06afaa9962"
   },
   "outputs": [],
   "source": [
    "fig = plt.figure(figsize=(20, 18))\n",
    "ax = fig.add_subplot(111, projection=\"3d\")\n",
    "ax.view_init(elev=70, azim=-70)\n",
    "ax.scatter(points[:, 0], points[:, 1], zs=-points[:, 2], c=sdf, cmap=\"hot_r\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "F6Wmzwyx5CrG"
   },
   "outputs": [],
   "source": [
    "points, sdf = sample_sdf_near_surface(bunny_trimesh, number_of_points=5000)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 1000
    },
    "colab_type": "code",
    "id": "ywtzOoFS5QaV",
    "outputId": "8563b586-8d29-449d-8864-7cc467e3b3bc"
   },
   "outputs": [],
   "source": [
    "fig = plt.figure(figsize=(20, 18))\n",
    "ax = fig.add_subplot(111, projection=\"3d\")\n",
    "ax.view_init(elev=70, azim=-70)\n",
    "ax.scatter(points[:, 0], points[:, 1], zs=-points[:, 2], c=sdf, cmap=\"hot_r\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "VwPfl3xhZysT"
   },
   "source": [
    "# Облако точек"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "QADjrqMD56LR"
   },
   "source": [
    "Облако точек модели bunny.obj, полученное семплированием с помощью алгоритма из pytorch3d."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 751
    },
    "colab_type": "code",
    "id": "tIZgYQkoMoLW",
    "outputId": "564a90ba-6cd7-46a6-cdf2-30b3700aaf02"
   },
   "outputs": [],
   "source": [
    "points = sample_points_from_meshes(bunny_mesh, 5000)\n",
    "points = points.clone().detach().cpu().squeeze().unbind(1)\n",
    "plot_pointcloud(points, elev=190, azim=150, title='Bunny pytotch3d mesh obj')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "2WoFTT5P6Y74"
   },
   "source": [
    "Облако точек модели sphere.obj, полученное семплированием с помощью алгоритма из pytorch3d."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 751
    },
    "colab_type": "code",
    "id": "vTNUMhf66aZC",
    "outputId": "6d4303fe-52e8-4cf0-bd15-1aedea4c9e56"
   },
   "outputs": [],
   "source": [
    "points = sample_points_from_meshes(sphere_mesh, 3000)\n",
    "points = points.clone().detach().cpu().squeeze().unbind(1)\n",
    "plot_pointcloud(points, elev=190, azim=130, title='Sphere pytotch3d mesh obj')"
   ]
  }
 ],
 "metadata": {
  "accelerator": "GPU",
  "colab": {
   "collapsed_sections": [],
   "name": "3dml-habr-phygitalism.ipynb",
   "provenance": [],
   "toc_visible": true
  },
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}