task3.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# <a id=\"#\">Task 3 – Mean, median, and Gaussian filtering</a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---\n",
    "\n",
    "## Linear filtering by convolution (mean filter)\n",
    "\n",
    "**<a id=\"#\">Task 3.1.1.</a>** Load the image `data/astronaut.png` and show it:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**<a id=\"#\">Task 3.1.2.</a>** Implement a mean filter:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def meanfilter(img_in, size):\n",
    "    result = numpy.zeros(img_in.shape)\n",
    "    n = (size - 1) // 2\n",
    "    for p in numpy.ndindex(img_in.shape):\n",
    "        neighborhood_sum = 0\n",
    "        for q in numpy.ndindex(size, size):\n",
    "            ...\n",
    "    return result"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**<a id=\"#\">Task 3.1.3.</a>** Test your implementation from above by using the function `meanfilter` for the previously loaded image and show the result:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**<a id=\"#\">Task 3.1.4.</a>** Compare your result for filter size `5` with the correct result image `data/astronaut_meanfilter5.png`:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---\n",
    "\n",
    "## Non-linear filtering\n",
    "\n",
    "**<a id=\"#\">Task 3.2.1.</a>** Implement a median filter:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def medianfilter(img_in, size):\n",
    "    result = numpy.zeros(img_in.shape)\n",
    "    ...\n",
    "    return result"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**<a id=\"#\">Task 3.2.2.</a>** First, test your implementation from above by using the function `medianfilter` for the previously loaded image and show the result:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Then, compare your result for filter size `5` with the correct result image `data/astronaut_medianfilter5.png`:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---\n",
    "\n",
    "## Using pre-implemented filters\n",
    "\n",
    "**<a id=\"#\">Task 3.3.1.</a>** Load the package `scipy.ndimage`:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**<a id=\"#\">Task 3.3.2 (a).</a>** Apply a **mean filter** from `scipy.ndimage` to the loaded image:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**<a id=\"#\">Task 3.3.2 (b).</a>** Apply a **median filter** from `scipy.ndimage` to the loaded image:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**<a id=\"#\">Task 3.3.2 (c).</a>** Apply a **Gaussian filter** from `scipy.ndimage` to the loaded image:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**<a id=\"#\">Task 3.3.3.</a>** Compare the results obtained using the functions from `scipy.ndimage` with those you obtained using your own implementations. What are the main differences? Do you have an explanation?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---\n",
    "\n",
    "## Slicing and benchmarking\n",
    "\n",
    "**<a id=\"#\">Task 3.4.2.</a>** Implement either a mean filter or a median filter using only a single `for`-loop:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def fastfilter(img_in, size):\n",
    "    ..."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**<a id=\"#\">Task 3.4.3.</a>** First, test your implementation from above by using the function `fastfilter` for the previously loaded image and show the result:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Then, compare your result to those obtained using your implementation for either `meanfilter` or `medianfilter`, respectively:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**<a id=\"#\">Task 3.4.4.</a>** Benchmark the code run times:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**<a id=\"#\">Task 3.4.5.</a>** Document your observations regarding the benchmarking results and try to think of an explanation:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.12.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}