3D_seed_pixelflow_demo.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "GO0vsHOgGwgL"
      },
      "source": [
        "# Automated Extraction of 3D Seed Phenotype Data using a pre-trained `StarDist` model, `Scivision`, and `Pixelflow`"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "joKES48-khez"
      },
      "source": [
        "### This notebook demonstrates how to use pixelflow to extract measurements of segmented objects from 3D imagery. This demo will use 3D x-ray computed tomography images of Oilseed rape (*Brassica napus*) seedpods detected and segmented using an adapted StarDist model as described in [Corcoran et al. 2023](https://www.frontiersin.org/articles/10.3389/fpls.2023.1120182/full).\n",
        "\n",
        "### Example images required to run this notebook can be found [here](https://zenodo.org/records/8355920)\n",
        "\n",
        "### R code used to run valve sorting in this notebook is available from the [Scivision gallery github page](https://github.com/scivision-gallery/pixelflow_seed_demo/tree/main) seed (`'seedpod_valve_lowess_single.R'`)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Tq16E6b_Gi2g"
      },
      "source": [
        "## 1. Install and load relevant packages"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {
        "id": "ItdBJ2m6k0LQ"
      },
      "outputs": [],
      "source": [
        "from pixelflow import pixelflow\n",
        "\n",
        "import matplotlib.pyplot as plt\n",
        "import numpy as np\n",
        "from tifffile import imread\n",
        "from pathlib import Path\n",
        "\n",
        "from glob import glob\n",
        "import re\n",
        "\n",
        "from scivision import default_catalog, load_pretrained_model"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "au80VydpG8fK"
      },
      "source": [
        "## 2. Find `StarDist Seed` model in `Scivision` model catalog"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 2,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "8V_KIYw7HBf3",
        "outputId": "4b34d08f-5833-492c-b296-b44cf2d3f577"
      },
      "outputs": [
        {
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>name</th>\n",
              "      <th>description</th>\n",
              "      <th>tasks</th>\n",
              "      <th>pkg_url</th>\n",
              "      <th>url</th>\n",
              "      <th>scivision_usable</th>\n",
              "      <th>institution</th>\n",
              "      <th>tags</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>stardist</td>\n",
              "      <td>Single class object detection and segementatio...</td>\n",
              "      <td>(TaskEnum.object_detection, TaskEnum.segmentat...</td>\n",
              "      <td>git+https://github.com/stardist/stardist.git@m...</td>\n",
              "      <td>https://github.com/stardist/stardist</td>\n",
              "      <td>False</td>\n",
              "      <td>(epfl,)</td>\n",
              "      <td>(2D, 3D, optical-microscopy, xray, microtomogr...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>PlantCV</td>\n",
              "      <td>Open-source image analysis software package ta...</td>\n",
              "      <td>(TaskEnum.thresholding, TaskEnum.object_detect...</td>\n",
              "      <td>git+https://github.com/danforthcenter/plantcv@...</td>\n",
              "      <td>https://github.com/danforthcenter/plantcv</td>\n",
              "      <td>False</td>\n",
              "      <td>(danforthcenter,)</td>\n",
              "      <td>(2D, hyperspectral, multispectral, near-infrar...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>greenotyper</td>\n",
              "      <td>Greenotyper is an image analysis tool for larg...</td>\n",
              "      <td>(TaskEnum.segmentation)</td>\n",
              "      <td>git+https://github.com/MarniTausen/Greenotyper...</td>\n",
              "      <td>https://github.com/MarniTausen/Greenotyper</td>\n",
              "      <td>False</td>\n",
              "      <td>(MarniTausen,)</td>\n",
              "      <td>(2D, multispectral, plant-phenotyping, climate...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>mapreader-plant</td>\n",
              "      <td>automated detection of plant patches in images...</td>\n",
              "      <td>(TaskEnum.object_detection, TaskEnum.classific...</td>\n",
              "      <td>git+https://github.com/alan-turing-institute/m...</td>\n",
              "      <td>https://github.com/alan-turing-institute/mapre...</td>\n",
              "      <td>True</td>\n",
              "      <td>(alan-turing-institute,)</td>\n",
              "      <td>(2D, plant, phenotype, rgb, biology, agriculture)</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>resnet50-plankton</td>\n",
              "      <td>automated classification of plankton images co...</td>\n",
              "      <td>(TaskEnum.classificiation)</td>\n",
              "      <td>git+https://github.com/alan-turing-institute/p...</td>\n",
              "      <td>https://github.com/alan-turing-institute/plank...</td>\n",
              "      <td>True</td>\n",
              "      <td>(alan-turing-institute, cefas, plankton-analyt...</td>\n",
              "      <td>(2D, plankton, ecology, environmental-science)</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>vedge-detector</td>\n",
              "      <td>automated detection of coastal vegetation edge...</td>\n",
              "      <td>(TaskEnum.segmentation)</td>\n",
              "      <td>git+https://github.com/MartinSJRogers/VEdge_De...</td>\n",
              "      <td>https://github.com/MartinSJRogers/VEdge_Detect...</td>\n",
              "      <td>True</td>\n",
              "      <td>(cambridge-university, birkbeck-university)</td>\n",
              "      <td>(2D, satellite, remote-sensing, ecology, envir...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>image-classifiers</td>\n",
              "      <td>Scivision plugin for the image-classifiers pac...</td>\n",
              "      <td>(TaskEnum.classificiation)</td>\n",
              "      <td>git+https://github.com/alan-turing-institute/s...</td>\n",
              "      <td>https://github.com/alan-turing-institute/scivi...</td>\n",
              "      <td>True</td>\n",
              "      <td>(alan-turing-institute,)</td>\n",
              "      <td>(classification, 2D, image)</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>huggingface-classifiers</td>\n",
              "      <td>Selection of image classifiers taken from http...</td>\n",
              "      <td>(TaskEnum.classificiation)</td>\n",
              "      <td>git+https://github.com/alan-turing-institute/s...</td>\n",
              "      <td>https://github.com/alan-turing-institute/scivi...</td>\n",
              "      <td>True</td>\n",
              "      <td>(alan-turing-institute, huggingface)</td>\n",
              "      <td>(classification, 2D, image)</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>huggingface-segmentation</td>\n",
              "      <td>Selection of segmentation algorithms taken fro...</td>\n",
              "      <td>(TaskEnum.segmentation)</td>\n",
              "      <td>git+https://github.com/alan-turing-institute/s...</td>\n",
              "      <td>https://github.com/alan-turing-institute/scivi...</td>\n",
              "      <td>True</td>\n",
              "      <td>(alan-turing-institute, huggingface)</td>\n",
              "      <td>()</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>huggingface-object-detection</td>\n",
              "      <td>Selection of object detection algorithms taken...</td>\n",
              "      <td>(TaskEnum.object_detection)</td>\n",
              "      <td>git+https://github.com/alan-turing-institute/s...</td>\n",
              "      <td>https://github.com/alan-turing-institute/scivi...</td>\n",
              "      <td>True</td>\n",
              "      <td>(alan-turing-institute, huggingface)</td>\n",
              "      <td>()</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10</th>\n",
              "      <td>detectreeRGB-forest</td>\n",
              "      <td>Detect changes in tall tree height with machin...</td>\n",
              "      <td>(TaskEnum.object_detection, TaskEnum.segmentat...</td>\n",
              "      <td>git+https://github.com/shmh40/detectreeRGB-tre...</td>\n",
              "      <td>https://github.com/shmh40/detectreeRGB-treecro...</td>\n",
              "      <td>True</td>\n",
              "      <td>(cambridge-university,)</td>\n",
              "      <td>(2D, satellite, remote-sensing, ecology, envir...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>11</th>\n",
              "      <td>flower-classification-model</td>\n",
              "      <td>Identify flower species</td>\n",
              "      <td>(TaskEnum.classificiation)</td>\n",
              "      <td>git+https://github.com/alan-turing-institute/f...</td>\n",
              "      <td>https://github.com/alan-turing-institute/flowe...</td>\n",
              "      <td>True</td>\n",
              "      <td>(alan-turing-institute,)</td>\n",
              "      <td>(flowers, ecology)</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>12</th>\n",
              "      <td>butterfly-classification-model</td>\n",
              "      <td>Identify butterfly species</td>\n",
              "      <td>(TaskEnum.classificiation)</td>\n",
              "      <td>git+https://github.com/alan-turing-institute/b...</td>\n",
              "      <td>https://github.com/alan-turing-institute/butte...</td>\n",
              "      <td>True</td>\n",
              "      <td>(alan-turing-institute,)</td>\n",
              "      <td>(butterflies, insects, animals, ecology)</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>13</th>\n",
              "      <td>Affinity-VAE</td>\n",
              "      <td>A framework for automatic clustering and class...</td>\n",
              "      <td>(TaskEnum.classificiation)</td>\n",
              "      <td>https://github.com/alan-turing-institute/affin...</td>\n",
              "      <td>https://github.com/alan-turing-institute/affin...</td>\n",
              "      <td>False</td>\n",
              "      <td>(The Alan Turing Institute, Rutherford Appleto...</td>\n",
              "      <td>(Protein, Tomography, Cryo-EM, Shape regularis...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>14</th>\n",
              "      <td>btrack</td>\n",
              "      <td>btrack is a Python library for multi object tr...</td>\n",
              "      <td>()</td>\n",
              "      <td>btrack</td>\n",
              "      <td>https://github.com/quantumjot/btrack</td>\n",
              "      <td>False</td>\n",
              "      <td>(The Alan Turing Institute, University College...</td>\n",
              "      <td>(cell tracking, bio imaging , optimisation, ba...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>15</th>\n",
              "      <td>StarDist Seed</td>\n",
              "      <td>Pre-trained StarDist model for detection and s...</td>\n",
              "      <td>(TaskEnum.object_detection, TaskEnum.segmentat...</td>\n",
              "      <td>stardist_seed_pod</td>\n",
              "      <td>https://raw.githubusercontent.com/alan-turing-...</td>\n",
              "      <td>True</td>\n",
              "      <td>(The Alan Turing Institute, Rothamsted Research)</td>\n",
              "      <td>(2D, 3D, plant-phenotyping, agriculture, rgb, ...</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                              name  \\\n",
              "0                         stardist   \n",
              "1                          PlantCV   \n",
              "2                      greenotyper   \n",
              "3                  mapreader-plant   \n",
              "4                resnet50-plankton   \n",
              "5                   vedge-detector   \n",
              "6                image-classifiers   \n",
              "7          huggingface-classifiers   \n",
              "8         huggingface-segmentation   \n",
              "9     huggingface-object-detection   \n",
              "10             detectreeRGB-forest   \n",
              "11     flower-classification-model   \n",
              "12  butterfly-classification-model   \n",
              "13                    Affinity-VAE   \n",
              "14                          btrack   \n",
              "15                   StarDist Seed   \n",
              "\n",
              "                                          description  \\\n",
              "0   Single class object detection and segementatio...   \n",
              "1   Open-source image analysis software package ta...   \n",
              "2   Greenotyper is an image analysis tool for larg...   \n",
              "3   automated detection of plant patches in images...   \n",
              "4   automated classification of plankton images co...   \n",
              "5   automated detection of coastal vegetation edge...   \n",
              "6   Scivision plugin for the image-classifiers pac...   \n",
              "7   Selection of image classifiers taken from http...   \n",
              "8   Selection of segmentation algorithms taken fro...   \n",
              "9   Selection of object detection algorithms taken...   \n",
              "10  Detect changes in tall tree height with machin...   \n",
              "11                            Identify flower species   \n",
              "12                         Identify butterfly species   \n",
              "13  A framework for automatic clustering and class...   \n",
              "14  btrack is a Python library for multi object tr...   \n",
              "15  Pre-trained StarDist model for detection and s...   \n",
              "\n",
              "                                                tasks  \\\n",
              "0   (TaskEnum.object_detection, TaskEnum.segmentat...   \n",
              "1   (TaskEnum.thresholding, TaskEnum.object_detect...   \n",
              "2                             (TaskEnum.segmentation)   \n",
              "3   (TaskEnum.object_detection, TaskEnum.classific...   \n",
              "4                          (TaskEnum.classificiation)   \n",
              "5                             (TaskEnum.segmentation)   \n",
              "6                          (TaskEnum.classificiation)   \n",
              "7                          (TaskEnum.classificiation)   \n",
              "8                             (TaskEnum.segmentation)   \n",
              "9                         (TaskEnum.object_detection)   \n",
              "10  (TaskEnum.object_detection, TaskEnum.segmentat...   \n",
              "11                         (TaskEnum.classificiation)   \n",
              "12                         (TaskEnum.classificiation)   \n",
              "13                         (TaskEnum.classificiation)   \n",
              "14                                                 ()   \n",
              "15  (TaskEnum.object_detection, TaskEnum.segmentat...   \n",
              "\n",
              "                                              pkg_url  \\\n",
              "0   git+https://github.com/stardist/stardist.git@m...   \n",
              "1   git+https://github.com/danforthcenter/plantcv@...   \n",
              "2   git+https://github.com/MarniTausen/Greenotyper...   \n",
              "3   git+https://github.com/alan-turing-institute/m...   \n",
              "4   git+https://github.com/alan-turing-institute/p...   \n",
              "5   git+https://github.com/MartinSJRogers/VEdge_De...   \n",
              "6   git+https://github.com/alan-turing-institute/s...   \n",
              "7   git+https://github.com/alan-turing-institute/s...   \n",
              "8   git+https://github.com/alan-turing-institute/s...   \n",
              "9   git+https://github.com/alan-turing-institute/s...   \n",
              "10  git+https://github.com/shmh40/detectreeRGB-tre...   \n",
              "11  git+https://github.com/alan-turing-institute/f...   \n",
              "12  git+https://github.com/alan-turing-institute/b...   \n",
              "13  https://github.com/alan-turing-institute/affin...   \n",
              "14                                             btrack   \n",
              "15                                  stardist_seed_pod   \n",
              "\n",
              "                                                  url  scivision_usable  \\\n",
              "0                https://github.com/stardist/stardist             False   \n",
              "1           https://github.com/danforthcenter/plantcv             False   \n",
              "2          https://github.com/MarniTausen/Greenotyper             False   \n",
              "3   https://github.com/alan-turing-institute/mapre...              True   \n",
              "4   https://github.com/alan-turing-institute/plank...              True   \n",
              "5   https://github.com/MartinSJRogers/VEdge_Detect...              True   \n",
              "6   https://github.com/alan-turing-institute/scivi...              True   \n",
              "7   https://github.com/alan-turing-institute/scivi...              True   \n",
              "8   https://github.com/alan-turing-institute/scivi...              True   \n",
              "9   https://github.com/alan-turing-institute/scivi...              True   \n",
              "10  https://github.com/shmh40/detectreeRGB-treecro...              True   \n",
              "11  https://github.com/alan-turing-institute/flowe...              True   \n",
              "12  https://github.com/alan-turing-institute/butte...              True   \n",
              "13  https://github.com/alan-turing-institute/affin...             False   \n",
              "14               https://github.com/quantumjot/btrack             False   \n",
              "15  https://raw.githubusercontent.com/alan-turing-...              True   \n",
              "\n",
              "                                          institution  \\\n",
              "0                                             (epfl,)   \n",
              "1                                   (danforthcenter,)   \n",
              "2                                      (MarniTausen,)   \n",
              "3                            (alan-turing-institute,)   \n",
              "4   (alan-turing-institute, cefas, plankton-analyt...   \n",
              "5         (cambridge-university, birkbeck-university)   \n",
              "6                            (alan-turing-institute,)   \n",
              "7                (alan-turing-institute, huggingface)   \n",
              "8                (alan-turing-institute, huggingface)   \n",
              "9                (alan-turing-institute, huggingface)   \n",
              "10                            (cambridge-university,)   \n",
              "11                           (alan-turing-institute,)   \n",
              "12                           (alan-turing-institute,)   \n",
              "13  (The Alan Turing Institute, Rutherford Appleto...   \n",
              "14  (The Alan Turing Institute, University College...   \n",
              "15   (The Alan Turing Institute, Rothamsted Research)   \n",
              "\n",
              "                                                 tags  \n",
              "0   (2D, 3D, optical-microscopy, xray, microtomogr...  \n",
              "1   (2D, hyperspectral, multispectral, near-infrar...  \n",
              "2   (2D, multispectral, plant-phenotyping, climate...  \n",
              "3   (2D, plant, phenotype, rgb, biology, agriculture)  \n",
              "4      (2D, plankton, ecology, environmental-science)  \n",
              "5   (2D, satellite, remote-sensing, ecology, envir...  \n",
              "6                         (classification, 2D, image)  \n",
              "7                         (classification, 2D, image)  \n",
              "8                                                  ()  \n",
              "9                                                  ()  \n",
              "10  (2D, satellite, remote-sensing, ecology, envir...  \n",
              "11                                 (flowers, ecology)  \n",
              "12           (butterflies, insects, animals, ecology)  \n",
              "13  (Protein, Tomography, Cryo-EM, Shape regularis...  \n",
              "14  (cell tracking, bio imaging , optimisation, ba...  \n",
              "15  (2D, 3D, plant-phenotyping, agriculture, rgb, ...  "
            ]
          },
          "execution_count": 2,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "# load full scivision model catalog\n",
        "models_catalog = default_catalog.models.to_dataframe()\n",
        "models_catalog"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 3,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 52
        },
        "id": "ITolsBTvIBXP",
        "outputId": "94eb2b90-89d3-4819-9d8b-411def3602f5"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "FlexibleUrl('https://raw.githubusercontent.com/alan-turing-institute/stardist-seed-pod/main/.scivision/model.yml', scheme='https', host='raw.githubusercontent.com', tld='com', host_type='domain', path='/alan-turing-institute/stardist-seed-pod/main/.scivision/model.yml')"
            ]
          },
          "execution_count": 3,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "#get github repository link for the model we want to use (StarDist Seed)\n",
        "model_repo = models_catalog[models_catalog.name == \"StarDist Seed\"].url.item()\n",
        "model_repo"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "c05wj_wfIGn1"
      },
      "source": [
        "## 3. Load `StarDist Seed` model from `Scivision` model catalog"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 4,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "3l96TV3JIMeW",
        "outputId": "9a65ed1d-70b7-4f18-d5df-9c219ef6eeb9"
      },
      "outputs": [
        {
          "data": {
            "text/html": [
              "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"color: #7fbfbf; text-decoration-color: #7fbfbf\">[09:48:25] </span><span style=\"color: #000080; text-decoration-color: #000080\">INFO    </span> <span style=\"color: #808000; text-decoration-color: #808000\">Call</span>=<span style=\"color: #008000; text-decoration-color: #008000\">'load_pretrained_model'</span>, <span style=\"color: #808000; text-decoration-color: #808000\">module</span>=<span style=\"color: #008000; text-decoration-color: #008000\">'scivision.io.reader'</span>, <span style=\"color: #808000; text-decoration-color: #808000\">signature</span>=<span style=\"font-weight: bold\">(</span>path:        <a href=\"file:///Users/ifenton/opt/anaconda3/envs/test_pxsd4/lib/python3.10/site-packages/scivision/koala.py\" target=\"_blank\"><span style=\"color: #7f7f7f; text-decoration-color: #7f7f7f\">koala.py</span></a><span style=\"color: #7f7f7f; text-decoration-color: #7f7f7f\">:</span><a href=\"file:///Users/ifenton/opt/anaconda3/envs/test_pxsd4/lib/python3.10/site-packages/scivision/koala.py#49\" target=\"_blank\"><span style=\"color: #7f7f7f; text-decoration-color: #7f7f7f\">49</span></a>\n",
              "<span style=\"color: #7fbfbf; text-decoration-color: #7fbfbf\">           </span>         os.PathLike, branch: str = <span style=\"color: #008000; text-decoration-color: #008000\">'main'</span>, allow_install: bool = <span style=\"color: #ff0000; text-decoration-color: #ff0000; font-style: italic\">False</span>, model_selection:    <span style=\"color: #7f7f7f; text-decoration-color: #7f7f7f\">           </span>\n",
              "<span style=\"color: #7fbfbf; text-decoration-color: #7fbfbf\">           </span>         str = <span style=\"color: #008000; text-decoration-color: #008000\">'default'</span>, load_multiple: bool = <span style=\"color: #ff0000; text-decoration-color: #ff0000; font-style: italic\">False</span>, *args, **kwargs<span style=\"font-weight: bold\">)</span> -&gt;                   <span style=\"color: #7f7f7f; text-decoration-color: #7f7f7f\">           </span>\n",
              "<span style=\"color: #7fbfbf; text-decoration-color: #7fbfbf\">           </span>         scivision.io.wrapper.PretrainedModel, <span style=\"color: #808000; text-decoration-color: #808000\">kwargs_given</span>=<span style=\"color: #008000; text-decoration-color: #008000\">'model_selection, allow_install'</span> <span style=\"color: #7f7f7f; text-decoration-color: #7f7f7f\">           </span>\n",
              "</pre>\n"
            ],
            "text/plain": [
              "\u001b[2;36m[09:48:25]\u001b[0m\u001b[2;36m \u001b[0m\u001b[34mINFO    \u001b[0m \u001b[33mCall\u001b[0m=\u001b[32m'load_pretrained_model'\u001b[0m, \u001b[33mmodule\u001b[0m=\u001b[32m'scivision.io.reader'\u001b[0m, \u001b[33msignature\u001b[0m=\u001b[1m(\u001b[0mpath:        \u001b]8;id=32438;file:///Users/ifenton/opt/anaconda3/envs/test_pxsd4/lib/python3.10/site-packages/scivision/koala.py\u001b\\\u001b[2mkoala.py\u001b[0m\u001b]8;;\u001b\\\u001b[2m:\u001b[0m\u001b]8;id=876671;file:///Users/ifenton/opt/anaconda3/envs/test_pxsd4/lib/python3.10/site-packages/scivision/koala.py#49\u001b\\\u001b[2m49\u001b[0m\u001b]8;;\u001b\\\n",
              "\u001b[2;36m           \u001b[0m         os.PathLike, branch: str = \u001b[32m'main'\u001b[0m, allow_install: bool = \u001b[3;91mFalse\u001b[0m, model_selection:    \u001b[2m           \u001b[0m\n",
              "\u001b[2;36m           \u001b[0m         str = \u001b[32m'default'\u001b[0m, load_multiple: bool = \u001b[3;91mFalse\u001b[0m, *args, **kwargs\u001b[1m)\u001b[0m ->                   \u001b[2m           \u001b[0m\n",
              "\u001b[2;36m           \u001b[0m         scivision.io.wrapper.PretrainedModel, \u001b[33mkwargs_given\u001b[0m=\u001b[32m'model_selection, allow_install'\u001b[0m \u001b[2m           \u001b[0m\n"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "2024-02-06 09:48:27.867486: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  SSE4.1 SSE4.2\n",
            "To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Loading network weights from 'weights_best.h5'.\n",
            "Loading thresholds from 'thresholds.json'.\n",
            "Using default values: prob_thresh=0.7, nms_thresh=0.3.\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "2024-02-06 09:48:41.916766: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  SSE4.1 SSE4.2\n",
            "To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.\n"
          ]
        },
        {
          "data": {
            "text/plain": [
              "scivision.PretrainedModel( \n",
              "  module='stardist_seed_pod', \n",
              "  model='model_3d', \n",
              "  source='https://github.com/alan-turing-institute/stardist-seed-pod' \n",
              "  pipe='DataPipe(input=<Parameter \"X: numpy.ndarray\">, output=<Parameter \"image: numpy.ndarray\">)' \n",
              ")"
            ]
          },
          "execution_count": 4,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "#load pretrained seed model\n",
        "model = load_pretrained_model(model_repo, model_selection='model_3d', allow_install=True)\n",
        "model"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "JpTmdQTeITmW"
      },
      "source": [
        "# Extract seed size and shape metrics from a single image (one seed pod) with `Pixelflow`"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "7zZFlSO7IX3K"
      },
      "source": [
        "## 4. Load images"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 5,
      "metadata": {
        "id": "Ity8jCBok5NL"
      },
      "outputs": [],
      "source": [
        "#load image\n",
        "img = imread('C0007197.tif')\n",
        "pod_name = 'C0007197'"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2NdtGSMoIkYB"
      },
      "source": [
        "## 5. Predict location, size and shape of seeds with `StarDist Seed` model using `Scivision `core function"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 6,
      "metadata": {
        "id": "e3qZLjGThenj"
      },
      "outputs": [],
      "source": [
        "#detect and segment seeds in image\n",
        "pred = model.predict(img)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 7,
      "metadata": {
        "id": "4HcL3bR5hhV8"
      },
      "outputs": [],
      "source": [
        "#save seed labels created with Stardist Seed model\n",
        "np.save(Path('content/'+ pod_name + '_pred.npy'), pred)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "0xhMRjO-k52S"
      },
      "source": [
        "## 6. Calculate seed size, shape and arrangement metrics with `Pixelflow`"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 8,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 308
        },
        "id": "jk4vrj-yxaYD",
        "outputId": "a2b4d307-759f-44db-d7ec-6cea8d695597"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "<matplotlib.image.AxesImage at 0x1a6937d60>"
            ]
          },
          "execution_count": 8,
          "metadata": {},
          "output_type": "execute_result"
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 640x480 with 2 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "#load image and mask\n",
        "img = imread('C0007197.tif')\n",
        "mask = pred\n",
        "\n",
        "fig, ax = plt.subplots(1, 2)\n",
        "ax[0].imshow(img[img.shape[0]//2])\n",
        "ax[1].imshow(mask[mask.shape[0]//2])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 9,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "dvbc-tEGp9qB",
        "outputId": "513d8a08-47aa-4790-e699-9a5d15042c50"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "ZYX image detected\n"
          ]
        }
      ],
      "source": [
        "#create table of size and shape metrics\n",
        "features = pixelflow(\n",
        "    mask,\n",
        "    img,\n",
        "    spacing=(0.03,0.03,0.03),\n",
        "    features=('label', 'bbox', 'centroid', 'volume', 'surface_area', 'equivalent_diameter', 'major_axis_length', 'minor_axis_length', 'sphericity', 'solidity', 'image_intensity'),\n",
        ")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 10,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 697
        },
        "id": "fyxRqKPrqKKE",
        "outputId": "e1a7a6a1-36eb-4d17-f0fe-59c6e6a9df7c"
      },
      "outputs": [
        {
          "data": {
            "text/html": [
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>label</th>\n",
              "      <th>minor_axis_length</th>\n",
              "      <th>major_axis_length</th>\n",
              "      <th>equivalent_diameter</th>\n",
              "      <th>centroid-0</th>\n",
              "      <th>centroid-1</th>\n",
              "      <th>centroid-2</th>\n",
              "      <th>solidity</th>\n",
              "      <th>bbox-0</th>\n",
              "      <th>bbox-1</th>\n",
              "      <th>bbox-2</th>\n",
              "      <th>bbox-3</th>\n",
              "      <th>bbox-4</th>\n",
              "      <th>bbox-5</th>\n",
              "      <th>volume</th>\n",
              "      <th>sphericity</th>\n",
              "      <th>surface_area</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>1</td>\n",
              "      <td>0.981203</td>\n",
              "      <td>1.125500</td>\n",
              "      <td>1.033669</td>\n",
              "      <td>7.057080</td>\n",
              "      <td>9.182320</td>\n",
              "      <td>6.880791</td>\n",
              "      <td>0.944606</td>\n",
              "      <td>218</td>\n",
              "      <td>290</td>\n",
              "      <td>212</td>\n",
              "      <td>252</td>\n",
              "      <td>323</td>\n",
              "      <td>249</td>\n",
              "      <td>0.578286</td>\n",
              "      <td>0.840046</td>\n",
              "      <td>3.995854</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>2</td>\n",
              "      <td>0.967379</td>\n",
              "      <td>1.094070</td>\n",
              "      <td>1.021443</td>\n",
              "      <td>8.402966</td>\n",
              "      <td>8.423045</td>\n",
              "      <td>6.856169</td>\n",
              "      <td>0.948071</td>\n",
              "      <td>264</td>\n",
              "      <td>263</td>\n",
              "      <td>213</td>\n",
              "      <td>296</td>\n",
              "      <td>299</td>\n",
              "      <td>246</td>\n",
              "      <td>0.558009</td>\n",
              "      <td>0.897028</td>\n",
              "      <td>3.654030</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>3</td>\n",
              "      <td>0.931310</td>\n",
              "      <td>1.081708</td>\n",
              "      <td>0.993973</td>\n",
              "      <td>9.786055</td>\n",
              "      <td>9.224335</td>\n",
              "      <td>6.038981</td>\n",
              "      <td>0.942772</td>\n",
              "      <td>311</td>\n",
              "      <td>292</td>\n",
              "      <td>184</td>\n",
              "      <td>342</td>\n",
              "      <td>324</td>\n",
              "      <td>219</td>\n",
              "      <td>0.514188</td>\n",
              "      <td>0.881338</td>\n",
              "      <td>3.521730</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>4</td>\n",
              "      <td>0.993609</td>\n",
              "      <td>1.141612</td>\n",
              "      <td>1.068682</td>\n",
              "      <td>10.523322</td>\n",
              "      <td>8.308133</td>\n",
              "      <td>6.350884</td>\n",
              "      <td>0.946646</td>\n",
              "      <td>334</td>\n",
              "      <td>259</td>\n",
              "      <td>194</td>\n",
              "      <td>368</td>\n",
              "      <td>296</td>\n",
              "      <td>229</td>\n",
              "      <td>0.639063</td>\n",
              "      <td>0.857784</td>\n",
              "      <td>4.182815</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>5</td>\n",
              "      <td>0.972978</td>\n",
              "      <td>1.101306</td>\n",
              "      <td>1.022085</td>\n",
              "      <td>11.903140</td>\n",
              "      <td>8.292396</td>\n",
              "      <td>6.207130</td>\n",
              "      <td>0.944272</td>\n",
              "      <td>381</td>\n",
              "      <td>259</td>\n",
              "      <td>191</td>\n",
              "      <td>412</td>\n",
              "      <td>295</td>\n",
              "      <td>224</td>\n",
              "      <td>0.559062</td>\n",
              "      <td>0.883887</td>\n",
              "      <td>3.713021</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>6</td>\n",
              "      <td>1.039169</td>\n",
              "      <td>1.100054</td>\n",
              "      <td>1.064269</td>\n",
              "      <td>12.965118</td>\n",
              "      <td>9.242935</td>\n",
              "      <td>5.640875</td>\n",
              "      <td>0.947243</td>\n",
              "      <td>415</td>\n",
              "      <td>292</td>\n",
              "      <td>172</td>\n",
              "      <td>450</td>\n",
              "      <td>325</td>\n",
              "      <td>206</td>\n",
              "      <td>0.631179</td>\n",
              "      <td>0.904121</td>\n",
              "      <td>3.935736</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>7</td>\n",
              "      <td>0.998850</td>\n",
              "      <td>1.109968</td>\n",
              "      <td>1.049012</td>\n",
              "      <td>13.755882</td>\n",
              "      <td>8.356721</td>\n",
              "      <td>6.144574</td>\n",
              "      <td>0.946674</td>\n",
              "      <td>442</td>\n",
              "      <td>261</td>\n",
              "      <td>188</td>\n",
              "      <td>475</td>\n",
              "      <td>297</td>\n",
              "      <td>222</td>\n",
              "      <td>0.604422</td>\n",
              "      <td>0.891599</td>\n",
              "      <td>3.877405</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>8</td>\n",
              "      <td>0.996690</td>\n",
              "      <td>1.113522</td>\n",
              "      <td>1.040445</td>\n",
              "      <td>14.965326</td>\n",
              "      <td>9.286118</td>\n",
              "      <td>5.784761</td>\n",
              "      <td>0.943947</td>\n",
              "      <td>482</td>\n",
              "      <td>293</td>\n",
              "      <td>176</td>\n",
              "      <td>516</td>\n",
              "      <td>326</td>\n",
              "      <td>211</td>\n",
              "      <td>0.589734</td>\n",
              "      <td>0.876820</td>\n",
              "      <td>3.878623</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>9</td>\n",
              "      <td>1.023502</td>\n",
              "      <td>1.117775</td>\n",
              "      <td>1.057196</td>\n",
              "      <td>16.262782</td>\n",
              "      <td>9.277510</td>\n",
              "      <td>5.758335</td>\n",
              "      <td>0.945024</td>\n",
              "      <td>525</td>\n",
              "      <td>292</td>\n",
              "      <td>175</td>\n",
              "      <td>559</td>\n",
              "      <td>326</td>\n",
              "      <td>210</td>\n",
              "      <td>0.618678</td>\n",
              "      <td>0.897796</td>\n",
              "      <td>3.910957</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>10</td>\n",
              "      <td>1.011778</td>\n",
              "      <td>1.101909</td>\n",
              "      <td>1.058210</td>\n",
              "      <td>18.252815</td>\n",
              "      <td>8.195342</td>\n",
              "      <td>6.873644</td>\n",
              "      <td>0.949508</td>\n",
              "      <td>592</td>\n",
              "      <td>256</td>\n",
              "      <td>213</td>\n",
              "      <td>626</td>\n",
              "      <td>291</td>\n",
              "      <td>247</td>\n",
              "      <td>0.620460</td>\n",
              "      <td>0.912699</td>\n",
              "      <td>3.854480</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10</th>\n",
              "      <td>11</td>\n",
              "      <td>1.024113</td>\n",
              "      <td>1.076558</td>\n",
              "      <td>1.046852</td>\n",
              "      <td>18.387724</td>\n",
              "      <td>9.049965</td>\n",
              "      <td>6.177095</td>\n",
              "      <td>0.944834</td>\n",
              "      <td>596</td>\n",
              "      <td>285</td>\n",
              "      <td>190</td>\n",
              "      <td>630</td>\n",
              "      <td>319</td>\n",
              "      <td>224</td>\n",
              "      <td>0.600696</td>\n",
              "      <td>0.890054</td>\n",
              "      <td>3.868160</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>11</th>\n",
              "      <td>12</td>\n",
              "      <td>0.964471</td>\n",
              "      <td>1.104019</td>\n",
              "      <td>1.020652</td>\n",
              "      <td>20.204123</td>\n",
              "      <td>8.948896</td>\n",
              "      <td>6.644708</td>\n",
              "      <td>0.945739</td>\n",
              "      <td>659</td>\n",
              "      <td>282</td>\n",
              "      <td>204</td>\n",
              "      <td>690</td>\n",
              "      <td>316</td>\n",
              "      <td>240</td>\n",
              "      <td>0.556713</td>\n",
              "      <td>0.876403</td>\n",
              "      <td>3.734229</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>12</th>\n",
              "      <td>13</td>\n",
              "      <td>0.991399</td>\n",
              "      <td>1.146471</td>\n",
              "      <td>1.052763</td>\n",
              "      <td>21.099917</td>\n",
              "      <td>7.866746</td>\n",
              "      <td>7.754831</td>\n",
              "      <td>0.943381</td>\n",
              "      <td>687</td>\n",
              "      <td>246</td>\n",
              "      <td>240</td>\n",
              "      <td>720</td>\n",
              "      <td>279</td>\n",
              "      <td>278</td>\n",
              "      <td>0.610929</td>\n",
              "      <td>0.865637</td>\n",
              "      <td>4.022311</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>13</th>\n",
              "      <td>14</td>\n",
              "      <td>1.031616</td>\n",
              "      <td>1.095727</td>\n",
              "      <td>1.060492</td>\n",
              "      <td>21.757217</td>\n",
              "      <td>8.631629</td>\n",
              "      <td>7.164398</td>\n",
              "      <td>0.947522</td>\n",
              "      <td>709</td>\n",
              "      <td>271</td>\n",
              "      <td>222</td>\n",
              "      <td>742</td>\n",
              "      <td>306</td>\n",
              "      <td>258</td>\n",
              "      <td>0.624483</td>\n",
              "      <td>0.906223</td>\n",
              "      <td>3.898788</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>14</th>\n",
              "      <td>15</td>\n",
              "      <td>0.992368</td>\n",
              "      <td>1.135102</td>\n",
              "      <td>1.070860</td>\n",
              "      <td>22.622915</td>\n",
              "      <td>7.536766</td>\n",
              "      <td>8.240096</td>\n",
              "      <td>0.946916</td>\n",
              "      <td>738</td>\n",
              "      <td>234</td>\n",
              "      <td>256</td>\n",
              "      <td>770</td>\n",
              "      <td>270</td>\n",
              "      <td>294</td>\n",
              "      <td>0.642978</td>\n",
              "      <td>0.871074</td>\n",
              "      <td>4.135804</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>15</th>\n",
              "      <td>16</td>\n",
              "      <td>0.979231</td>\n",
              "      <td>1.123466</td>\n",
              "      <td>1.048356</td>\n",
              "      <td>24.057848</td>\n",
              "      <td>7.360396</td>\n",
              "      <td>8.585115</td>\n",
              "      <td>0.946018</td>\n",
              "      <td>786</td>\n",
              "      <td>227</td>\n",
              "      <td>268</td>\n",
              "      <td>818</td>\n",
              "      <td>264</td>\n",
              "      <td>304</td>\n",
              "      <td>0.603288</td>\n",
              "      <td>0.871888</td>\n",
              "      <td>3.960105</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>16</th>\n",
              "      <td>17</td>\n",
              "      <td>0.939268</td>\n",
              "      <td>1.075626</td>\n",
              "      <td>1.006022</td>\n",
              "      <td>25.219111</td>\n",
              "      <td>7.943015</td>\n",
              "      <td>8.195695</td>\n",
              "      <td>0.944918</td>\n",
              "      <td>825</td>\n",
              "      <td>248</td>\n",
              "      <td>257</td>\n",
              "      <td>856</td>\n",
              "      <td>283</td>\n",
              "      <td>291</td>\n",
              "      <td>0.533115</td>\n",
              "      <td>0.869452</td>\n",
              "      <td>3.656950</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>17</th>\n",
              "      <td>18</td>\n",
              "      <td>1.013690</td>\n",
              "      <td>1.096644</td>\n",
              "      <td>1.058716</td>\n",
              "      <td>26.068852</td>\n",
              "      <td>6.910379</td>\n",
              "      <td>9.311235</td>\n",
              "      <td>0.946453</td>\n",
              "      <td>852</td>\n",
              "      <td>213</td>\n",
              "      <td>293</td>\n",
              "      <td>886</td>\n",
              "      <td>248</td>\n",
              "      <td>328</td>\n",
              "      <td>0.621351</td>\n",
              "      <td>0.870457</td>\n",
              "      <td>4.045401</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>18</th>\n",
              "      <td>19</td>\n",
              "      <td>0.962153</td>\n",
              "      <td>1.141531</td>\n",
              "      <td>1.064875</td>\n",
              "      <td>26.413046</td>\n",
              "      <td>7.689681</td>\n",
              "      <td>8.734708</td>\n",
              "      <td>0.947021</td>\n",
              "      <td>865</td>\n",
              "      <td>238</td>\n",
              "      <td>274</td>\n",
              "      <td>897</td>\n",
              "      <td>275</td>\n",
              "      <td>310</td>\n",
              "      <td>0.632259</td>\n",
              "      <td>0.863547</td>\n",
              "      <td>4.125355</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>19</th>\n",
              "      <td>20</td>\n",
              "      <td>0.848878</td>\n",
              "      <td>1.124899</td>\n",
              "      <td>0.959285</td>\n",
              "      <td>28.189449</td>\n",
              "      <td>6.516726</td>\n",
              "      <td>9.806767</td>\n",
              "      <td>0.940088</td>\n",
              "      <td>926</td>\n",
              "      <td>200</td>\n",
              "      <td>311</td>\n",
              "      <td>955</td>\n",
              "      <td>236</td>\n",
              "      <td>343</td>\n",
              "      <td>0.462213</td>\n",
              "      <td>0.827153</td>\n",
              "      <td>3.495100</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>"
            ],
            "text/plain": [
              "    label  minor_axis_length  major_axis_length  equivalent_diameter  \\\n",
              "0       1           0.981203           1.125500             1.033669   \n",
              "1       2           0.967379           1.094070             1.021443   \n",
              "2       3           0.931310           1.081708             0.993973   \n",
              "3       4           0.993609           1.141612             1.068682   \n",
              "4       5           0.972978           1.101306             1.022085   \n",
              "5       6           1.039169           1.100054             1.064269   \n",
              "6       7           0.998850           1.109968             1.049012   \n",
              "7       8           0.996690           1.113522             1.040445   \n",
              "8       9           1.023502           1.117775             1.057196   \n",
              "9      10           1.011778           1.101909             1.058210   \n",
              "10     11           1.024113           1.076558             1.046852   \n",
              "11     12           0.964471           1.104019             1.020652   \n",
              "12     13           0.991399           1.146471             1.052763   \n",
              "13     14           1.031616           1.095727             1.060492   \n",
              "14     15           0.992368           1.135102             1.070860   \n",
              "15     16           0.979231           1.123466             1.048356   \n",
              "16     17           0.939268           1.075626             1.006022   \n",
              "17     18           1.013690           1.096644             1.058716   \n",
              "18     19           0.962153           1.141531             1.064875   \n",
              "19     20           0.848878           1.124899             0.959285   \n",
              "\n",
              "    centroid-0  centroid-1  centroid-2  solidity  bbox-0  bbox-1  bbox-2  \\\n",
              "0     7.057080    9.182320    6.880791  0.944606     218     290     212   \n",
              "1     8.402966    8.423045    6.856169  0.948071     264     263     213   \n",
              "2     9.786055    9.224335    6.038981  0.942772     311     292     184   \n",
              "3    10.523322    8.308133    6.350884  0.946646     334     259     194   \n",
              "4    11.903140    8.292396    6.207130  0.944272     381     259     191   \n",
              "5    12.965118    9.242935    5.640875  0.947243     415     292     172   \n",
              "6    13.755882    8.356721    6.144574  0.946674     442     261     188   \n",
              "7    14.965326    9.286118    5.784761  0.943947     482     293     176   \n",
              "8    16.262782    9.277510    5.758335  0.945024     525     292     175   \n",
              "9    18.252815    8.195342    6.873644  0.949508     592     256     213   \n",
              "10   18.387724    9.049965    6.177095  0.944834     596     285     190   \n",
              "11   20.204123    8.948896    6.644708  0.945739     659     282     204   \n",
              "12   21.099917    7.866746    7.754831  0.943381     687     246     240   \n",
              "13   21.757217    8.631629    7.164398  0.947522     709     271     222   \n",
              "14   22.622915    7.536766    8.240096  0.946916     738     234     256   \n",
              "15   24.057848    7.360396    8.585115  0.946018     786     227     268   \n",
              "16   25.219111    7.943015    8.195695  0.944918     825     248     257   \n",
              "17   26.068852    6.910379    9.311235  0.946453     852     213     293   \n",
              "18   26.413046    7.689681    8.734708  0.947021     865     238     274   \n",
              "19   28.189449    6.516726    9.806767  0.940088     926     200     311   \n",
              "\n",
              "    bbox-3  bbox-4  bbox-5    volume  sphericity  surface_area  \n",
              "0      252     323     249  0.578286    0.840046      3.995854  \n",
              "1      296     299     246  0.558009    0.897028      3.654030  \n",
              "2      342     324     219  0.514188    0.881338      3.521730  \n",
              "3      368     296     229  0.639063    0.857784      4.182815  \n",
              "4      412     295     224  0.559062    0.883887      3.713021  \n",
              "5      450     325     206  0.631179    0.904121      3.935736  \n",
              "6      475     297     222  0.604422    0.891599      3.877405  \n",
              "7      516     326     211  0.589734    0.876820      3.878623  \n",
              "8      559     326     210  0.618678    0.897796      3.910957  \n",
              "9      626     291     247  0.620460    0.912699      3.854480  \n",
              "10     630     319     224  0.600696    0.890054      3.868160  \n",
              "11     690     316     240  0.556713    0.876403      3.734229  \n",
              "12     720     279     278  0.610929    0.865637      4.022311  \n",
              "13     742     306     258  0.624483    0.906223      3.898788  \n",
              "14     770     270     294  0.642978    0.871074      4.135804  \n",
              "15     818     264     304  0.603288    0.871888      3.960105  \n",
              "16     856     283     291  0.533115    0.869452      3.656950  \n",
              "17     886     248     328  0.621351    0.870457      4.045401  \n",
              "18     897     275     310  0.632259    0.863547      4.125355  \n",
              "19     955     236     343  0.462213    0.827153      3.495100  "
            ]
          },
          "execution_count": 10,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "#show table of size and shape metrics in millimetres\n",
        "features"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 11,
      "metadata": {
        "id": "lDxhrIM5a_iT"
      },
      "outputs": [],
      "source": [
        "#output csv with measurements\n",
        "features.to_csv('C0007197.csv', header=True, index=False)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_UQCDJ-qii-j"
      },
      "source": [
        "## 7. Save segmented images of each individual seed with `Pixelflow`"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 12,
      "metadata": {
        "id": "ORnlKSC7ilR7"
      },
      "outputs": [],
      "source": [
        "#make list of all seeds in mask, shift backwards by one to match index values\n",
        "X = list(features.features.label)\n",
        "for i in range(len(X)):\n",
        "    X[i] -= 1"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 13,
      "metadata": {
        "id": "p0kHzhYOipvu"
      },
      "outputs": [],
      "source": [
        "#save segmented images of all seeds within a seed pod\n",
        "for file in X:\n",
        "    np.save(pod_name + '_seed_' + str(features.features.label[file]) + '.npy', features.image_intensity[file])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 14,
      "metadata": {
        "id": "cTRSQkhxiutf"
      },
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "#load single seed image\n",
        "seed_image = np.load('C0007197_seed_1.npy')\n",
        "\n",
        "fig = plt.imshow(seed_image[:, seed_image.shape[1]//2])"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "lbEifhAHi2Ra"
      },
      "source": [
        "# Extract seed size and shape metrics and single seed images from multiple images (one seed pod per image) with `Pixelflow`"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 15,
      "metadata": {
        "id": "XaUXhlIMi4GP"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "['content/C0007158.tif',\n",
              " 'content/C0007159.tif',\n",
              " 'content/C0007198.tif',\n",
              " 'content/C0007205.tif',\n",
              " 'content/C0007226.tif',\n",
              " 'content/C0007239.tif']"
            ]
          },
          "execution_count": 15,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "#load all input images\n",
        "img_list = sorted(glob('content/*.tif'))\n",
        "img_list"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 16,
      "metadata": {
        "id": "DqtHzT1UjAi8"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "ZYX image detected\n",
            "ZYX image detected\n",
            "ZYX image detected\n",
            "ZYX image detected\n",
            "ZYX image detected\n",
            "ZYX image detected\n"
          ]
        }
      ],
      "source": [
        "for file in img_list:\n",
        "  #load image and mask\n",
        "  img = imread(file)\n",
        "  pod_name = re.split(\"[/.]\", file)[-2]\n",
        "  pred = model.predict(img)\n",
        "  np.save(Path('content/'+ pod_name + '_pred.npy'), pred)\n",
        "  mask = pred\n",
        "\n",
        "  #create table of size and shape metrics\n",
        "  features = pixelflow(\n",
        "    mask,\n",
        "    img,\n",
        "    spacing=(0.03,0.03,0.03),\n",
        "    features=('label', 'bbox', 'centroid', 'volume', 'surface_area', 'equivalent_diameter', 'major_axis_length', 'minor_axis_length', 'sphericity', 'solidity', 'image_intensity'),\n",
        "  )\n",
        "  #output csv with measurements\n",
        "  features.to_csv(pod_name + '.csv', header=True, index=False)\n",
        "\n",
        "  #make list of all seeds in mask, shift backwards by one to match index values\n",
        "  X = list(features.features.label)\n",
        "  for i in range(len(X)):\n",
        "      X[i] -= 1\n",
        "\n",
        "  #save segmented images of all seeds within a seed pod\n",
        "  for file in X:\n",
        "      np.save(pod_name + '_seed_' + str(features.features.label[file]) + '.npy', features.image_intensity[file])"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "mmEnFvIXjKD3"
      },
      "source": [
        "## 8. Sort detected seeds into 2 seeds pod valves, and calculated related measurments (order in valve from seed pod beak to pedicel, space between seeds in each valve) with `'seedpod_valve_lowess_single.R'`"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 17,
      "metadata": {},
      "outputs": [],
      "source": [
        "import rpy2.robjects as robjects\n",
        "from rpy2.robjects import pandas2ri\n",
        "from rpy2.robjects import conversion, default_converter\n",
        "import rpy2.robjects.packages as rpackages\n",
        "from rpy2.robjects.vectors import StrVector"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 18,
      "metadata": {
        "id": "2AuRF7GDjQRC"
      },
      "outputs": [],
      "source": [
        "# R packages to install\n",
        "with conversion.localconverter(default_converter):\n",
        "  utils = rpackages.importr('utils')\n",
        "  utils.chooseCRANmirror(ind=1)\n",
        "  packnames = ('gplots', 'tidyverse')\n",
        "  # Selectively install what needs to be install.\n",
        "  names_to_install = [x for x in packnames if not rpackages.isinstalled(x)]\n",
        "  if len(names_to_install) > 0:\n",
        "      utils.install_packages(StrVector(names_to_install))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 19,
      "metadata": {
        "id": "Eu596uoCjTE0"
      },
      "outputs": [],
      "source": [
        "# Defining the R script and loading the instance in Python\n",
        "r = robjects.r"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 20,
      "metadata": {
        "id": "uJzAuXx_jWma"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "ZYX image detected\n",
            "── Attaching core tidyverse packages ──────────────────────── tidyverse 2.0.0 ──\n",
            "✔ dplyr     1.1.4     ✔ readr     2.1.5\n",
            "✔ forcats   1.0.0     ✔ stringr   1.5.1\n",
            "✔ ggplot2   3.4.4     ✔ tibble    3.2.1\n",
            "✔ lubridate 1.9.3     ✔ tidyr     1.3.1\n",
            "✔ purrr     1.0.2     \n",
            "── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──\n",
            "✖ dplyr::filter() masks stats::filter()\n",
            "✖ dplyr::lag()    masks stats::lag()\n",
            "ℹ Use the conflicted package (<http://conflicted.r-lib.org/>) to force all conflicts to become errors\n"
          ]
        },
        {
          "data": {
            "text/html": [
              "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"color: #7fbfbf; text-decoration-color: #7fbfbf\">[10:31:20] </span><span style=\"color: #800000; text-decoration-color: #800000\">WARNING </span> R<span style=\"font-weight: bold\">[</span>write to console<span style=\"font-weight: bold\">]</span>:                                                           <a href=\"file:///Users/ifenton/opt/anaconda3/envs/test_pxsd4/lib/python3.10/site-packages/rpy2/rinterface_lib/callbacks.py\" target=\"_blank\"><span style=\"color: #7f7f7f; text-decoration-color: #7f7f7f\">callbacks.py</span></a><span style=\"color: #7f7f7f; text-decoration-color: #7f7f7f\">:</span><a href=\"file:///Users/ifenton/opt/anaconda3/envs/test_pxsd4/lib/python3.10/site-packages/rpy2/rinterface_lib/callbacks.py#124\" target=\"_blank\"><span style=\"color: #7f7f7f; text-decoration-color: #7f7f7f\">124</span></a>\n",
              "<span style=\"color: #7fbfbf; text-decoration-color: #7fbfbf\">           </span>         Attaching package: ‘gplots’                                                    <span style=\"color: #7f7f7f; text-decoration-color: #7f7f7f\">                </span>\n",
              "<span style=\"color: #7fbfbf; text-decoration-color: #7fbfbf\">           </span>                                                                                        <span style=\"color: #7f7f7f; text-decoration-color: #7f7f7f\">                </span>\n",
              "<span style=\"color: #7fbfbf; text-decoration-color: #7fbfbf\">           </span>                                                                                        <span style=\"color: #7f7f7f; text-decoration-color: #7f7f7f\">                </span>\n",
              "</pre>\n"
            ],
            "text/plain": [
              "\u001b[2;36m[10:31:20]\u001b[0m\u001b[2;36m \u001b[0m\u001b[31mWARNING \u001b[0m R\u001b[1m[\u001b[0mwrite to console\u001b[1m]\u001b[0m:                                                           \u001b]8;id=840231;file:///Users/ifenton/opt/anaconda3/envs/test_pxsd4/lib/python3.10/site-packages/rpy2/rinterface_lib/callbacks.py\u001b\\\u001b[2mcallbacks.py\u001b[0m\u001b]8;;\u001b\\\u001b[2m:\u001b[0m\u001b]8;id=926700;file:///Users/ifenton/opt/anaconda3/envs/test_pxsd4/lib/python3.10/site-packages/rpy2/rinterface_lib/callbacks.py#124\u001b\\\u001b[2m124\u001b[0m\u001b]8;;\u001b\\\n",
              "\u001b[2;36m           \u001b[0m         Attaching package: ‘gplots’                                                    \u001b[2m                \u001b[0m\n",
              "\u001b[2;36m           \u001b[0m                                                                                        \u001b[2m                \u001b[0m\n",
              "\u001b[2;36m           \u001b[0m                                                                                        \u001b[2m                \u001b[0m\n"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "data": {
            "text/html": [
              "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"color: #7fbfbf; text-decoration-color: #7fbfbf\">           </span><span style=\"color: #800000; text-decoration-color: #800000\">WARNING </span> R<span style=\"font-weight: bold\">[</span>write to console<span style=\"font-weight: bold\">]</span>: The following object is masked from ‘package:stats’:      <a href=\"file:///Users/ifenton/opt/anaconda3/envs/test_pxsd4/lib/python3.10/site-packages/rpy2/rinterface_lib/callbacks.py\" target=\"_blank\"><span style=\"color: #7f7f7f; text-decoration-color: #7f7f7f\">callbacks.py</span></a><span style=\"color: #7f7f7f; text-decoration-color: #7f7f7f\">:</span><a href=\"file:///Users/ifenton/opt/anaconda3/envs/test_pxsd4/lib/python3.10/site-packages/rpy2/rinterface_lib/callbacks.py#124\" target=\"_blank\"><span style=\"color: #7f7f7f; text-decoration-color: #7f7f7f\">124</span></a>\n",
              "<span style=\"color: #7fbfbf; text-decoration-color: #7fbfbf\">           </span>                                                                                        <span style=\"color: #7f7f7f; text-decoration-color: #7f7f7f\">                </span>\n",
              "<span style=\"color: #7fbfbf; text-decoration-color: #7fbfbf\">           </span>             lowess                                                                     <span style=\"color: #7f7f7f; text-decoration-color: #7f7f7f\">                </span>\n",
              "<span style=\"color: #7fbfbf; text-decoration-color: #7fbfbf\">           </span>                                                                                        <span style=\"color: #7f7f7f; text-decoration-color: #7f7f7f\">                </span>\n",
              "<span style=\"color: #7fbfbf; text-decoration-color: #7fbfbf\">           </span>                                                                                        <span style=\"color: #7f7f7f; text-decoration-color: #7f7f7f\">                </span>\n",
              "</pre>\n"
            ],
            "text/plain": [
              "\u001b[2;36m          \u001b[0m\u001b[2;36m \u001b[0m\u001b[31mWARNING \u001b[0m R\u001b[1m[\u001b[0mwrite to console\u001b[1m]\u001b[0m: The following object is masked from ‘package:stats’:      \u001b]8;id=787429;file:///Users/ifenton/opt/anaconda3/envs/test_pxsd4/lib/python3.10/site-packages/rpy2/rinterface_lib/callbacks.py\u001b\\\u001b[2mcallbacks.py\u001b[0m\u001b]8;;\u001b\\\u001b[2m:\u001b[0m\u001b]8;id=409067;file:///Users/ifenton/opt/anaconda3/envs/test_pxsd4/lib/python3.10/site-packages/rpy2/rinterface_lib/callbacks.py#124\u001b\\\u001b[2m124\u001b[0m\u001b]8;;\u001b\\\n",
              "\u001b[2;36m           \u001b[0m                                                                                        \u001b[2m                \u001b[0m\n",
              "\u001b[2;36m           \u001b[0m             lowess                                                                     \u001b[2m                \u001b[0m\n",
              "\u001b[2;36m           \u001b[0m                                                                                        \u001b[2m                \u001b[0m\n",
              "\u001b[2;36m           \u001b[0m                                                                                        \u001b[2m                \u001b[0m\n"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "ZYX image detected\n",
            "ZYX image detected\n",
            "ZYX image detected\n",
            "ZYX image detected\n",
            "ZYX image detected\n"
          ]
        }
      ],
      "source": [
        "for file in img_list:\n",
        "  # Calculate the image data in pixelflow\n",
        "  img2 = imread(file)\n",
        "  pod_name = re.split(\"[/.]\", file)[-2]\n",
        "  mask2 = np.load(Path('content/'+ pod_name + '_pred.npy'))\n",
        "  features2 = pixelflow(mask2, img2, features=('label', 'bbox', 'centroid', 'volume', 'surface_area', 'equivalent_diameter', 'major_axis_length', 'minor_axis_length', 'sphericity', 'solidity', 'image_intensity'))\n",
        "  # Loading the function we have defined in R.\n",
        "  with conversion.localconverter(default_converter):\n",
        "      r['source']('seedpod_2D_valve_lowess_single.R')\n",
        "      seedpod_2D_single_r = robjects.globalenv['seedpod_2D_single']\n",
        "\n",
        "  #converting dataframe into r object for passing into r function\n",
        "  with (robjects.default_converter + pandas2ri.converter).context():\n",
        "    r_from_pd_df = robjects.conversion.get_conversion().py2rpy(features2.features)\n",
        "\n",
        "  #Invoking the R function and getting the result\n",
        "  with conversion.localconverter(default_converter):\n",
        "      df_result_r = seedpod_2D_single_r(r_from_pd_df, \"plot_\" + pod_name + \".png\")\n",
        "\n",
        "  #Converting it back to a pandas dataframe\n",
        "  with (robjects.default_converter + pandas2ri.converter).context():\n",
        "    df_result = robjects.conversion.get_conversion().rpy2py(df_result_r)"
      ]
    }
  ],
  "metadata": {
    "colab": {
      "provenance": []
    },
    "kernelspec": {
      "display_name": "Python 3",
      "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": "3.10.8"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}