{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"```{admonition} Information\n",
"__Section__: Image classification \n",
"__Goal__: Understand a way to perform image classification. \n",
"__Time needed__: 20 min \n",
"__Prerequisites__: Introduction about machine learning experiments\n",
"```"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Image classification"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Build the dataset\n",
"\n",
"In the previous section, we saw how to decompose an image into its HOG descriptors. This allows us to reduce the number of features that will be used to perform the classification.\n",
"\n",
"Those HOG descriptors can now be used as any normal set of attributes, just the same way we did in the previous chapter with AIS data. We only need to create a dataset, with one entry for each image, one attribute for each HOG descriptor, and a class attribute. As an example, a dataset for animal classification could look like that:\n",
"\n",
"\n",
"\n",
"For this to work fine, we need each image to have the same number of HOG descriptor. We saw in the previous section that the number of HOG descriptors depends on two parameters:\n",
"- the parameters chosen in the HOG function: we decide on the number of pixels to group together for the HOG descriptors. The more pixels are in the same group, the less HOG descriptors we get.\n",
"- the size of the image: similarly as for the parameters of the function, the bigger the image, the more HOG descriptors we get.\n",
"\n",
"In our case, it is therefore important for each image to have the same size (same number of pixels) and the same parameters for the HOG descriptors."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"With that in mind, we can now build our dataset by importing the images with their class and calculating the HOG descriptors for each of them. We are in possession of a .csv file ``2-images.csv`` containing the path and the class of each image. Then, we only need to loop over the images to calculate their HOG descriptor and put everything in a Pandas DataFrame."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"```{toggle} Advanced level\n",
"First, we read the .csv file.\n",
"```"
]
},
{
"cell_type": "code",
"execution_count": 1,
"metadata": {
"tags": [
"hide-input"
]
},
"outputs": [
{
"data": {
"text/html": [
"
\n",
"\n",
"
\n",
" \n",
"
\n",
"
\n",
"
image_path
\n",
"
label
\n",
"
\n",
" \n",
" \n",
"
\n",
"
0
\n",
"
./data/1.jpg
\n",
"
cat
\n",
"
\n",
"
\n",
"
1
\n",
"
./data/2.jpg
\n",
"
cat
\n",
"
\n",
"
\n",
"
2
\n",
"
./data/3.jpg
\n",
"
cat
\n",
"
\n",
"
\n",
"
3
\n",
"
./data/4.jpg
\n",
"
cat
\n",
"
\n",
"
\n",
"
4
\n",
"
./data/5.jpg
\n",
"
cat
\n",
"
\n",
" \n",
"
\n",
"
"
],
"text/plain": [
" image_path label\n",
"0 ./data/1.jpg cat\n",
"1 ./data/2.jpg cat\n",
"2 ./data/3.jpg cat\n",
"3 ./data/4.jpg cat\n",
"4 ./data/5.jpg cat"
]
},
"execution_count": 1,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"import pandas as pd\n",
"\n",
"images = pd.read_csv('./2-images.csv')\n",
"\n",
"images.head()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"```{toggle} Advanced level\n",
"Now, we can loop on each image path to read the image itself, calculate its HOG descriptors and put everything in a Dataframe for classification. We keep the read image in the DataFrame to be able to plot it later if needed.\n",
"```"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {
"tags": [
"hide-input"
]
},
"outputs": [
{
"data": {
"text/html": [
"
\n",
"\n",
"
\n",
" \n",
"
\n",
"
\n",
"
image_path
\n",
"
label
\n",
"
hog_features
\n",
"
\n",
" \n",
" \n",
"
\n",
"
0
\n",
"
./data/1.jpg
\n",
"
cat
\n",
"
[0.10489434136776081, 0.03000765876042029, 0.1...
\n",
"
\n",
"
\n",
"
1
\n",
"
./data/2.jpg
\n",
"
cat
\n",
"
[0.06511460912671653, 0.036009570336997296, 0....
\n",
"
\n",
"
\n",
"
2
\n",
"
./data/3.jpg
\n",
"
cat
\n",
"
[0.0645168443208118, 0.012273334354864062, 0.0...
\n",
"
\n",
"
\n",
"
3
\n",
"
./data/4.jpg
\n",
"
cat
\n",
"
[0.15030760450655362, 0.04317479024740703, 0.0...
\n",
"
\n",
"
\n",
"
4
\n",
"
./data/5.jpg
\n",
"
cat
\n",
"
[0.04515955710031582, 0.023032858411843354, 0....
\n",
"
\n",
" \n",
"
\n",
"
"
],
"text/plain": [
" image_path label hog_features\n",
"0 ./data/1.jpg cat [0.10489434136776081, 0.03000765876042029, 0.1...\n",
"1 ./data/2.jpg cat [0.06511460912671653, 0.036009570336997296, 0....\n",
"2 ./data/3.jpg cat [0.0645168443208118, 0.012273334354864062, 0.0...\n",
"3 ./data/4.jpg cat [0.15030760450655362, 0.04317479024740703, 0.0...\n",
"4 ./data/5.jpg cat [0.04515955710031582, 0.023032858411843354, 0...."
]
},
"execution_count": 2,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"from skimage import io\n",
"from skimage.feature import hog\n",
"\n",
"hog_features = []\n",
"for file in images['image_path']:\n",
" hog_features.append(hog(io.imread(file), # the image read\n",
" orientations = 8,\n",
" pixels_per_cell = (40, 40),\n",
" visualize = False)\n",
" )\n",
"\n",
"images = images.assign(hog_features = hog_features)\n",
"\n",
"images.head()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"```{toggle} Advanced level\n",
"In this example, all the HOG descriptors are stored as a list in a single attribute. This is not a problem for classification, we will unpack them when doing the classification to treat each of them as a single attribute.\n",
"```"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Perform the classification"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Now that we have the dataset, we can do the classification exactly how we are doing it usually."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"```{toggle} Advanced level\n",
"We choose to use a Random Forest classifier.\n",
"```"
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {
"tags": [
"hide-input"
]
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Accuracy: 0.25\n"
]
}
],
"source": [
"import numpy as np\n",
"from sklearn.ensemble import RandomForestClassifier\n",
"from sklearn.metrics import accuracy_score\n",
"from sklearn.model_selection import train_test_split\n",
"\n",
"train, test, y_train, y_test = train_test_split(images[['image_path', 'hog_features']], # we keep the attribute 'image_path' to\n",
" #be able to access the image to check the classification if needed\n",
" images['label'],\n",
" test_size = 0.2,\n",
" random_state = 0)\n",
"\n",
"x_train = np.stack(train['hog_features'].values)\n",
"x_test = np.stack(test['hog_features'].values)\n",
" \n",
"random_forest = RandomForestClassifier(n_estimators = 10, max_depth = 7, random_state = 0)\n",
"random_forest.fit(x_train, y_train.values)\n",
"predictions = random_forest.predict(x_test)\n",
" \n",
"accuracy = accuracy_score(predictions, y_test)\n",
"\n",
"print('Accuracy: ' + str(accuracy))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"After performing the classification, we can have a better look at the results to see which images were correctly or wrongly classified by our model:"
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {
"tags": [
"hide-input"
]
},
"outputs": [
{
"data": {
"image/png": "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\n",
"text/plain": [
""
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"import matplotlib.pyplot as plt\n",
"\n",
"i = 0\n",
"l = len(test)\n",
"for index, row in test.iterrows():\n",
" plt.subplot(1, l, i+1)\n",
" plt.imshow(io.imread(row['image_path']))\n",
" plt.title(str(predictions[i]))\n",
" i = i + 1\n",
"plt.show()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"In that case, we can see that the classification is very bad. This is very likely due to the very small size of the dataset.\n",
"\n",
"Indeed, the dataset contains 19 images, but each image contains 7200 HOG descriptors. The amount of images in the dataset is way too small."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create functions"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"```{toggle} Advanced level\n",
"To automatize the creation of the dataset, the classification of the images and the representation of the results, we can write a few functions that we can reuse with any other image dataset.\n",
"```"
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {
"tags": [
"hide-input"
]
},
"outputs": [],
"source": [
"def create_hog(images):\n",
" # Takes a list of images, returns a list of the hog descriptors\n",
" \n",
" from skimage.feature import hog\n",
"\n",
" hog_features = []\n",
" for image in images:\n",
" hog_features.append(hog(image,\n",
" orientations = 8,\n",
" pixels_per_cell = (40, 40),\n",
" visualize = False)\n",
" )\n",
" \n",
" return hog_features"
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {
"tags": [
"hide-input"
]
},
"outputs": [],
"source": [
"def classify_images(df, x, y, random_state = 0):\n",
" # From a df, performs image classification and returns the results\n",
" # Takes random_state as an input to allow different reproducible results\n",
" \n",
" from sklearn.ensemble import RandomForestClassifier\n",
" from sklearn.model_selection import train_test_split\n",
"\n",
" train, test, y_train, y_test = train_test_split(df[x], # we keep the attribute 'image_path' to\n",
" #be able to access the image to check the classification if needed\n",
" df[y],\n",
" test_size = 0.2,\n",
" random_state = random_state)\n",
"\n",
" x_train = np.stack(train['hog_features'].values)\n",
" x_test = np.stack(test['hog_features'].values)\n",
"\n",
" random_forest = RandomForestClassifier(n_estimators = 10, max_depth = 7, random_state = 0)\n",
" random_forest.fit(x_train, y_train.values)\n",
" predictions = random_forest.predict(x_test)\n",
" \n",
" return predictions, y_test, test"
]
},
{
"cell_type": "code",
"execution_count": 7,
"metadata": {
"tags": [
"hide-input"
]
},
"outputs": [],
"source": [
"def print_results(predictions, test):\n",
" \n",
" import matplotlib.pyplot as plt\n",
"\n",
" i = 0\n",
" l = len(test)\n",
" for index, row in test.iterrows():\n",
" plt.subplot(1, l, i+1)\n",
" plt.imshow(io.imread(row['image_path']))\n",
" plt.title(str(predictions[i]))\n",
" i = i + 1\n",
" plt.show()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"```{toggle} Advanced level\n",
"Now, to perform the classification, we only have to run those functions as following:\n",
"```"
]
},
{
"cell_type": "code",
"execution_count": 10,
"metadata": {
"tags": [
"hide-input"
]
},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"C:\\Users\\Anna\\Anaconda3\\lib\\site-packages\\ipykernel_launcher.py:18: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples,), for example using ravel().\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Accuracy: 0.5\n"
]
},
{
"data": {
"image/png": "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\n",
"text/plain": [
""
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"import pandas as pd\n",
"from skimage import io\n",
"from sklearn.metrics import accuracy_score\n",
"\n",
"images = pd.read_csv('./2-images.csv')\n",
"\n",
"list_images = []\n",
"for image in images['image_path']:\n",
" list_images.append(io.imread(image))\n",
"\n",
"hog_features = create_hog(list_images)\n",
"images = images.assign(hog_features = hog_features)\n",
"\n",
"x = ['image_path', 'hog_features']\n",
"y = ['label']\n",
"pred, y_test, test = classify_images(images, x, y, 1)\n",
"\n",
"print('Accuracy: ' + str(accuracy_score(pred, y_test)))\n",
"print_results(pred, test)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"We can try different values for ``random_state`` to see different results. Replace the value for ``random_state`` to visualize different results."
]
},
{
"cell_type": "code",
"execution_count": 12,
"metadata": {
"tags": [
"hide-input",
"hide-output"
]
},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"C:\\Users\\Anna\\Anaconda3\\lib\\site-packages\\ipykernel_launcher.py:18: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples,), for example using ravel().\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Accuracy: 0.75\n"
]
},
{
"data": {
"image/png": "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\n",
"text/plain": [
""
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"x = ['image_path', 'hog_features']\n",
"y = ['label']\n",
"pred, y_test, test = classify_images(images, x, y, 2)\n",
"print('Accuracy: ' + str(accuracy_score(pred, y_test)))\n",
"print_results(pred, test)"
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {
"tags": [
"hide-input"
]
},
"outputs": [
{
"data": {
"application/vnd.jupyter.widget-view+json": {
"model_id": "cdd48ac6f58c415eab425bb5932c58e4",
"version_major": 2,
"version_minor": 0
},
"text/plain": [
"interactive(children=(IntText(value=0, description='random_state = '), Output()), _dom_classes=('widget-intera…"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/plain": [
""
]
},
"execution_count": 5,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# Beginner version\n",
"\n",
"import matplotlib.pyplot as plt\n",
"import ipywidgets as widgets\n",
"from ipywidgets import interact\n",
"\n",
"def create_hog(images):\n",
" # Takes a list of images, returns a list of the hog descriptors\n",
" \n",
" from skimage.feature import hog\n",
"\n",
" hog_features = []\n",
" for image in images:\n",
" hog_features.append(hog(image,\n",
" orientations = 8,\n",
" pixels_per_cell = (40, 40),\n",
" visualize = False)\n",
" )\n",
" \n",
" return hog_features\n",
"\n",
"def classify_images(df, x, y, random_state = 0):\n",
" # From a df, performs image classification and returns the results\n",
" # Takes random_state as an input to allow different reproducible results\n",
" \n",
" from sklearn.ensemble import RandomForestClassifier\n",
" from sklearn.model_selection import train_test_split\n",
"\n",
" train, test, y_train, y_test = train_test_split(df[x], # we keep the attribute 'image_path' to\n",
" #be able to access the image to check the classification if needed\n",
" df[y],\n",
" test_size = 0.2,\n",
" random_state = random_state)\n",
"\n",
" x_train = np.stack(train['hog_features'].values)\n",
" x_test = np.stack(test['hog_features'].values)\n",
"\n",
" random_forest = RandomForestClassifier(n_estimators = 10, max_depth = 7, random_state = 0)\n",
" random_forest.fit(x_train, y_train.values)\n",
" predictions = random_forest.predict(x_test)\n",
" \n",
" return predictions, y_test, test\n",
"\n",
"def print_results(predictions, test):\n",
" \n",
" import matplotlib.pyplot as plt\n",
" from skimage import io\n",
"\n",
" i = 0\n",
" l = len(test)\n",
" for index, row in test.iterrows():\n",
" plt.subplot(1, l, i+1)\n",
" plt.imshow(io.imread(row['image_path']))\n",
" plt.title(str(predictions[i]))\n",
" i = i + 1\n",
" plt.show()\n",
"\n",
"def classify(random_state):\n",
" import pandas as pd\n",
" from skimage import io\n",
" from sklearn.metrics import accuracy_score\n",
"\n",
" images = pd.read_csv('./2-images.csv')\n",
"\n",
" list_images = []\n",
" for image in images['image_path']:\n",
" list_images.append(io.imread(image))\n",
"\n",
" hog_features = create_hog(list_images)\n",
" images = images.assign(hog_features = hog_features)\n",
"\n",
" x = ['image_path', 'hog_features']\n",
" y = ['label']\n",
" pred, y_test, test = classify_images(images, x, y, random_state)\n",
"\n",
" print('Accuracy: ' + str(accuracy_score(pred, y_test)))\n",
" print_results(pred, test)\n",
"\n",
" \n",
"interact(classify,\n",
" random_state = widgets.IntText(value = 0,\n",
" description = 'random_state = ',\n",
" disabled = False))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Quiz"
]
},
{
"cell_type": "code",
"execution_count": 1,
"metadata": {
"tags": [
"hide-input"
]
},
"outputs": [
{
"data": {
"text/html": [
"\n",
" \n",
" "
],
"text/plain": [
""
]
},
"execution_count": 1,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"from IPython.display import IFrame\n",
"IFrame(\"https://blog.hoou.de/wp-admin/admin-ajax.php?action=h5p_embed&id=62\", \"760\", \"399\")"
]
}
],
"metadata": {
"celltoolbar": "Edit 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": "3.7.1"
},
"varInspector": {
"cols": {
"lenName": 16,
"lenType": 16,
"lenVar": 40
},
"kernels_config": {
"python": {
"delete_cmd_postfix": "",
"delete_cmd_prefix": "del ",
"library": "var_list.py",
"varRefreshCmd": "print(var_dic_list())"
},
"r": {
"delete_cmd_postfix": ") ",
"delete_cmd_prefix": "rm(",
"library": "var_list.r",
"varRefreshCmd": "cat(var_dic_list()) "
}
},
"types_to_exclude": [
"module",
"function",
"builtin_function_or_method",
"instance",
"_Feature"
],
"window_display": false
}
},
"nbformat": 4,
"nbformat_minor": 4
}