PyBulletEnvSetup.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "PyBulletEnvSetup.ipynb",
      "provenance": [],
      "authorship_tag": "ABX9TyPHQIcznb+xFt79rnko7tQ5",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/ksengin/TB-Networks/blob/master/PyBulletEnvSetup.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "PXMQOe_aCeC1",
        "outputId": "c60f67c2-533c-4c22-f3f7-57dbcf55f3bf"
      },
      "source": [
        "## Installing pybullet might take ~5 minutes\n",
        "%%time\n",
        "!pip install pybullet"
      ],
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Collecting pybullet\n",
            "  Downloading pybullet-3.1.7.tar.gz (79.0 MB)\n",
            "\u001b[K     |████████████████████████████████| 79.0 MB 18 kB/s \n",
            "\u001b[?25hBuilding wheels for collected packages: pybullet\n",
            "  Building wheel for pybullet (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "  Created wheel for pybullet: filename=pybullet-3.1.7-cp37-cp37m-linux_x86_64.whl size=89750825 sha256=26a92211a5645ada12badd3ad9cffe30c20cc5ea348c334b8dd8079baa847820\n",
            "  Stored in directory: /root/.cache/pip/wheels/70/1c/62/86c8b68885c24123d87c5392d6678aa2b68a1796c8113e1aa6\n",
            "Successfully built pybullet\n",
            "Installing collected packages: pybullet\n",
            "Successfully installed pybullet-3.1.7\n",
            "CPU times: user 2.59 s, sys: 334 ms, total: 2.92 s\n",
            "Wall time: 5min 35s\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "VHTdXMbpCo0g"
      },
      "source": [
        "import pybullet as p\n",
        "import pybullet_data as pd\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "from matplotlib import pylab\n",
        "%matplotlib inline"
      ],
      "execution_count": 2,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gHwwrjTTC10B"
      },
      "source": [
        "class Simulation:\n",
        "    \"\"\"Class containing functions for the simulation environment setup\n",
        "    \"\"\"\n",
        "    def __init__(self, gravity:bool=True):\n",
        "        p.resetSimulation()\n",
        "        p.configureDebugVisualizer(p.COV_ENABLE_GUI)\n",
        "        self.useFixedBase = True\n",
        "        self.flags = p.URDF_INITIALIZE_SAT_FEATURES\n",
        "\n",
        "        if gravity:\n",
        "            self.set_gravity()\n",
        "\n",
        "    def reset(self):\n",
        "        p.resetSimulation()\n",
        "\n",
        "    def set_gravity(self):\n",
        "        p.setGravity(0, 0, -9.8)\n",
        "    \n",
        "\n",
        "    def add_model(self, filepath: str, pos: list) -> int:\n",
        "        assert len(pos) == 3\n",
        "\n",
        "        obj_id = p.loadURDF(filepath, pos, flags=self.flags, useFixedBase=self.useFixedBase)\n",
        "\n",
        "        return obj_id\n"
      ],
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1diiM0qcCwq5"
      },
      "source": [
        "class Camera:\n",
        "    \"\"\"Class containing functions to compute camera parameters and capture images\n",
        "    \"\"\"\n",
        "\n",
        "    def __init__(self):\n",
        "        self.pixel_width = 640\n",
        "        self.pixel_height = 480\n",
        "\n",
        "        nearPlane = 0.01\n",
        "        farPlane = 100\n",
        "        fov = 60\n",
        "        aspect = self.pixel_width / self.pixel_height\n",
        "\n",
        "        self.projection_matrix = p.computeProjectionMatrixFOV(fov, aspect, nearPlane, farPlane)\n",
        "        \n",
        "        self.up_axis_index = 2 # 'z'\n",
        "\n",
        "        self.shadow = 0\n",
        "        self.light_direction = None # [1, 1, 1]\n",
        "\n",
        "    def get_intrinsics(self) -> tuple:\n",
        "        return self.projection_matrix\n",
        "\n",
        "    def get_view_matrix_from_euler(self,\n",
        "                            cam_target_pos, \n",
        "                            cam_distance, \n",
        "                            yaw, \n",
        "                            pitch, \n",
        "                            roll) -> tuple:\n",
        "\n",
        "        view_matrix = p.computeViewMatrixFromYawPitchRoll(cam_target_pos, cam_distance, yaw, pitch,\n",
        "                                                                roll, self.up_axis_index)\n",
        "        return view_matrix\n",
        "\n",
        "\n",
        "    def capture(self, view_matrix) -> tuple:\n",
        "        img_arr = p.getCameraImage(self.pixel_width, \n",
        "                                    self.pixel_height, \n",
        "                                    view_matrix, \n",
        "                                    self.projection_matrix, \n",
        "                                    shadow=self.shadow,\n",
        "                                    lightDirection=self.light_direction)\n",
        "\n",
        "        rgb = img_arr[2]\n",
        "        depth = img_arr[3]\n",
        "        semantic = img_arr[4]\n",
        "\n",
        "        return rgb, depth, semantic\n"
      ],
      "execution_count": 4,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 287
        },
        "id": "MFUgrWOUCype",
        "outputId": "528f20fc-374c-406e-dd84-8cc744f76d8c"
      },
      "source": [
        "p.connect(p.DIRECT)\n",
        "p.setAdditionalSearchPath(pd.getDataPath())\n",
        "\n",
        "plane_pos = [0,0,-0.625]\n",
        "table_pos = [0,0,-0.625]\n",
        "\n",
        "sim = Simulation()\n",
        "plane_id = sim.add_model(\"plane.urdf\", plane_pos)\n",
        "table_id = sim.add_model(\"table/table.urdf\", table_pos)\n",
        "arb_id = sim.add_model(\"xarm/xarm6_robot.urdf\", np.zeros(3).tolist())\n",
        "\n",
        "cam = Camera()\n",
        "\n",
        "cam_target_pos = [0, 0, 0]\n",
        "\n",
        "yaw = 0\n",
        "pitch = -10.0\n",
        "roll = 0\n",
        "cam_distance = 2.5\n",
        "\n",
        "viewmatrix = cam.get_view_matrix_from_euler(cam_target_pos, cam_distance, yaw, pitch, roll)\n",
        "rgb, depth, semantic = cam.capture(viewmatrix)\n",
        "\n",
        "plt.imshow(rgb)"
      ],
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.image.AxesImage at 0x7fb168dfcf90>"
            ]
          },
          "metadata": {},
          "execution_count": 6
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    }
  ]
}