notebooks/Beginner_PTV.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "iJjxB7QvvdsL",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "f909e701-6c40-4b6e-e2f0-9cc6354cd0df"
      },
      "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 imageio-ffmpeg\n",
            "  Downloading imageio_ffmpeg-0.4.7-py3-none-manylinux2010_x86_64.whl (26.9 MB)\n",
            "\u001b[K     |████████████████████████████████| 26.9 MB 1.6 MB/s \n",
            "\u001b[?25hInstalling collected packages: imageio-ffmpeg\n",
            "Successfully installed imageio-ffmpeg-0.4.7\n"
          ]
        }
      ],
      "source": [
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "import imageio\n",
        "from scipy.ndimage import label, find_objects, center_of_mass, binary_erosion, binary_dilation\n",
        "!pip install imageio-ffmpeg"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "!wget http://www.benjymarks.com/alert/red-ball.MOV"
      ],
      "metadata": {
        "id": "6Gra72_6xjSi",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "bef705d4-18a5-4b11-89f7-5ebc08a21fb5"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "--2022-09-29 12:13:36--  http://www.benjymarks.com/alert/red-ball.MOV\n",
            "Resolving www.benjymarks.com (www.benjymarks.com)... 65.254.227.224\n",
            "Connecting to www.benjymarks.com (www.benjymarks.com)|65.254.227.224|:80... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 23337193 (22M) [video/quicktime]\n",
            "Saving to: ‘red-ball.MOV’\n",
            "\n",
            "red-ball.MOV        100%[===================>]  22.26M  19.7MB/s    in 1.1s    \n",
            "\n",
            "2022-09-29 12:13:38 (19.7 MB/s) - ‘red-ball.MOV’ saved [23337193/23337193]\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "video = imageio.get_reader('red-ball.MOV')"
      ],
      "metadata": {
        "id": "ByzokkQlvjc8",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "9368b897-2800-420f-86d3-8c55a940597a"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "WARNING:imageio_ffmpeg:The frame size for reading (720, 1280) is different from the source frame size (1280, 720).\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "structuring_element = np.array([[0,0,0,0,1,0,0,0,0],\n",
        "                                [0,0,1,1,1,1,1,0,0],\n",
        "                                [0,1,1,1,1,1,1,1,0],\n",
        "                                [0,1,1,1,1,1,1,1,0],\n",
        "                                [1,1,1,1,1,1,1,1,1],\n",
        "                                [0,1,1,1,1,1,1,1,0],\n",
        "                                [0,1,1,1,1,1,1,1,0],\n",
        "                                [0,0,1,1,1,1,1,0,0],\n",
        "                                [0,0,0,0,1,0,0,0,0]], dtype=bool)\n",
        "\n",
        "t = 0 # counter for frame number\n",
        "trajectory = [] # list to store the centroid location in\n",
        "\n",
        "for image in video.iter_data():\n",
        "    \n",
        "    im_hsv = rgb_to_hsv(image) # Convert to hsv\n",
        "\n",
        "    just_red = ((im_hsv[:,:,0] > 0.9) + (im_hsv[:,:,0] < 0.1)) * (im_hsv[:,:,1] > 0.7)\n",
        "\n",
        "    eroded =  binary_erosion(just_red, structuring_element)\n",
        "    opened =  binary_dilation(eroded, structuring_element)\n",
        "\n",
        "    labelled_image, num_features = label(opened)\n",
        "\n",
        "    objects = find_objects(labelled_image)\n",
        "\n",
        "    for i in range(len(objects)):\n",
        "        c = center_of_mass(opened, labelled_image, [i+1])[0]\n",
        "        area = labelled_image[objects[i]].sum()\n",
        "        if area > 10000:\n",
        "            # print(f'Area of object {i+1} is {area} pixels')\n",
        "            trajectory.append(c) \n",
        "    \n",
        "    t += 1"
      ],
      "metadata": {
        "id": "lqiT5sTbvetN"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "plt.figure(figsize=[5,12])\n",
        "trajectory = np.array(trajectory)\n",
        "plt.plot(trajectory[:,1],trajectory[:,0])\n",
        "plt.gca().invert_yaxis()\n",
        "plt.xlabel('x')\n",
        "plt.ylabel('y')\n",
        "plt.gca().set_aspect('equal')\n",
        "plt.show()"
      ],
      "metadata": {
        "id": "_PkJaiSXocgp",
        "outputId": "a76f05e7-86ea-42d7-ca58-697259b4b5fe",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 714
        }
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 360x864 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "oUfZKapTvnZW"
      },
      "execution_count": null,
      "outputs": []
    }
  ]
}