added gaussian filter

image-analysis-session/00_preprocessing.ipynb

@@ -66,6 +66,52 @@
     "    - *Here we will do something similar to adaptive background subtraction when we do adaptive thresholding*"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Prerequisite: Load an image"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "%matplotlib inline\n",
+    "import imageio\n",
+    "import io\n",
+    "import numpy as np\n",
+    "import requests\n",
+    "import skimage\n",
+    "\n",
+    "from matplotlib import pyplot as plt\n",
+    "from matplotlib import cm"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "#load and display an image\n",
+    "image_file = (\n",
+    "    \"https://git.embl.de/grp-bio-it/image-analysis-training-resources/raw/\"\n",
+    "    \"1fc71747f0caddfd3e206b700b41f2bf69c07a49/image_data/xy_8bit__two_cells.tif\"\n",
+    ")\n",
+    "\n",
+    "buffer = io.BytesIO()\n",
+    "buffer.write(requests.get(image_file).content)\n",
+    "buffer.seek(0)\n",
+    "\n",
+    "image = imageio.imread(buffer, format='tif')\n",
+    "f = plt.figure()\n",
+    "s = f.add_subplot(111)\n",
+    "s.imshow(image, cmap=cm.gray)"
+   ]
+  },
   {
    "cell_type": "markdown",
    "metadata": {},
@@ -82,9 +128,38 @@
     "\n",
     "**How to choose the correct value of $\\sigma$?**\n",
     "\n",
-    "This depends a lot on your images, in particular on the pixel size. In general, the chosen $\\sigma$ should be large enough to blur out noise but small enough so the \"structures of interest\" do not get blurred too much. Usually, the best value for $\\sigma$ is simply found by trying out some different options and looking at the result. \n"
+    "This depends a lot on your images, in particular on the pixel size. In general, the chosen $\\sigma$ should be large enough to blur out noise but small enough so the \"structures of interest\" do not get blurred too much. Usually, the best value for $\\sigma$ is simply found by trying out some different options and looking at the result.\n",
+    "\n",
+    "\n",
+    "Useful links:\n",
+    "\n",
+    "* [scikit-image filters](https://scikit-image.org/docs/dev/api/skimage.filters.html#skimage.filters.gaussian)\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "sigma = 3\n",
+    "filtered_image = skimage.filters.gaussian(image, sigma=sigma)\n",
+    "f = plt.figure()\n",
+    "s = f.add_subplot(121)\n",
+    "s.imshow(image, cmap=cm.gray)\n",
+    "s.set_title(\"original image\")\n",
+    "s = f.add_subplot(122)\n",
+    "s.imshow(filtered_image, cmap=cm.gray)\n",
+    "s.set_title(f\"filtered image with $\\sigma$ = {sigma}\")"
    ]
   },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  },
   {
    "cell_type": "code",
    "execution_count": null,