Merge pull request #9 from BiAPoL/feature-extraction

Add feature extraction slides/exercises to the repo

Author: jo-mueller

05_feature_extraction/00_plotting_in_python.ipynb

@@ -0,0 +1,192 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "id": "6595df14-f25f-4f5a-b631-c1abbebaa0bb",
+   "metadata": {},
+   "source": [
+    "# Exercise 1\n",
+    "\n",
+    "In this exercise, we will try out different threshold methods and see how they perform on the same data and create a label image from the binarized image data.\n",
+    "\n",
+    "## Recap\n",
+    "\n",
+    "Thresholding is the process of separating the background and the foreground of an image by comparing the intensities of each image to a single value, the threshold value. Finding a \"good\" threshold value is therefore of key importance. \n",
+    "\n",
+    "The output of a thresholding operation is a binary image, consiting of 0s and 1s - which is equivalent to `True` and `False` in terms of boolean values. \n",
+    "\n",
+    "**Vocabulary:** Thresholding is a type of segmentation. Separating the image into two types of pixels (background and foreground) is a type of semantic segmentation, and is - since only two types of pixels exist in the output - referred to as a binarization operation.\n",
+    "\n",
+    "**How to code:** Remember - in Python you can compare objects (single numbers or images) to threshold values simply by using the `<` or `>` operator - make sure to put the output into a new variable!"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 6,
+   "id": "96bb8728-eaf9-48aa-985b-f4eb94ed5c7f",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from skimage import data, filters, measure\n",
+    "import matplotlib.pyplot as plt\n",
+    "import napari"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "id": "2ab114eb-c9a6-417c-be01-a59105cf0202",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "image = data.human_mitosis()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "19ae7d4c-2382-428a-8c06-f6908f4710a3",
+   "metadata": {},
+   "source": [
+    "As a first task, use the `plt.imshow()` function to display the image in the notebook:"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "bb85b7d4-5fec-48e6-9dce-883218196c80",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Code goes here"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "1b3bef6e-bac7-4355-8387-f5f403cb73ea",
+   "metadata": {},
+   "source": [
+    "Now, apply several of the thresholds from `skimage` to the image data! Choose one that you think is suitable.\n",
+    "Hint: You can find them under `filters.threshold_...` - use the `tab` key or the [documentation](https://scikit-image.org/docs/stable/api/skimage.filters.html?highlight=filters#module-skimage.filters) to find out how to use the implemented threshold functions."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "2168ef40-5913-4d44-815b-6b328262b2e0",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "binary_image = # code goes here"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "ee9147cf-4773-4a01-9da4-5f5867b35ed5",
+   "metadata": {},
+   "source": [
+    "Again, use `plt.imshow()` to visualize the results"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "23ed60a0-9cb6-48d7-949a-8d572f693f9c",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Code goes here"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "82f5717f-2fad-495f-84eb-f0ba543b0922",
+   "metadata": {},
+   "source": [
+    "Now we would like to perform connected-component analysis. Use the appropriate function from skimage (`measure.label()`) for this task!"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "58550efa-a032-417d-90d2-e009f760a8db",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "label_image = # code goes here"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "8e316ab9-10e9-489a-96af-fb98e65b7ac8",
+   "metadata": {},
+   "source": [
+    "## Visualization in Napari"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 5,
+   "id": "f8fa699b-8764-4b8f-8967-2da4a292cdb1",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "WARNING: DirectWrite: CreateFontFaceFromHDC() failed (Indicates an error in an input file such as a font file.) for QFontDef(Family=\"8514oem\", pointsize=12, pixelsize=20, styleHint=5, weight=50, stretch=100, hintingPreference=0) LOGFONT(\"8514oem\", lfWidth=0, lfHeight=-20) dpi=240\n"
+     ]
+    }
+   ],
+   "source": [
+    "viewer = napari.Viewer()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "d8b83da7-0e93-4922-9468-cc0735ae5ce6",
+   "metadata": {},
+   "source": [
+    "Use the `add_image()` and the `add_labels()` function to add `image`, `binary_image` and `label_image` to the viewer and create screenshot of the viewer with the `napari.utils.nbscreenshot()` function."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 8,
+   "id": "0f648bb4-ee2e-41ac-befa-33f54f9f3843",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Code goes here"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "b6614988-1613-4a37-b6b0-591f8c643f38",
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3 (ipykernel)",
+   "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": "3.9.12"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}