[go: up one dir, main page]
Scribd
Upload
7 views3 pages

Tutorial 4 Data Augumentation

This tutorial covers data augmentation in deep learning, emphasizing its importance for increasing dataset size and variability when collecting large amounts of data is impractical. It outlines the steps to implement data augmentation using Keras, including importing libraries, defining save locations, configuring augmentation parameters, and generating augmented images. The tutorial concludes with a task to create and save 40 augmented versions of a single JPEG image.

Uploaded by

Muhammad Abbas
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
7 views3 pages

Tutorial 4 Data Augumentation

This tutorial covers data augmentation in deep learning, emphasizing its importance for increasing dataset size and variability when collecting large amounts of data is impractical. It outlines the steps to implement data augmentation using Keras, including importing libraries, defining save locations, configuring augmentation parameters, and generating augmented images. The tutorial concludes with a task to create and save 40 augmented versions of a single JPEG image.

Uploaded by

Muhammad Abbas
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 3

Tutorial 4:

Data Augmentation
Objectives:
• Understand the Need for data augmentation
• Understand Operations in data augmentation
• Understand Data augmentation in Keras

Data augmentation is an integral process in deep learning, as in deep


learning we need large amounts of data and in some cases, it is not
feasible to collect thousands or millions of images, so data augmentation
comes to the rescue. It helps us to increase the size of the dataset and
introduce variability in the dataset.

Step 1 — Importing Necessary Libraries


First, we need to import the required libraries.

Step 2 Define the Folder Where Augmented Images


Will Be Saved
Now, we define the directory where the augmented images will be saved. If the
folder doesn't exist, we'll create it.
Step 3: Initialize the ImageDataGenerator with
Augmentation Parameters
Here, we configure the ImageDataGenerator class with the
augmentation techniques we want to apply to the images.

Rotation: This rotates the image within a given range.

Shear: Shear introduces geometric transformations, skewing the image.

Zoom: It zooms in and out randomly on the image.

Flip: Flips the image horizontally.

Brightness: Randomly adjusts the brightness of the image.

Step 4: Load the Image and Convert It to a NumPy Array

We now load a sample image from a file and convert it into a NumPy array, which
is the format Keras works with.
Step 5: Reshape the Image for Batch Processing

ImageDataGenerator expects a 4D array: (batch_size, height, width, channels).


Since we are working with a single image, we add an extra dimension.

Step 6: Generate and Save Augmented Images

We now create a loop that generates augmented images. Each image will be saved
in the specified directory, and optionally, we will display the augmented images
using matplotlib.

Step 7: Confirm the Saved Location of Augmented Images

Finally, we print out the folder location where the augmented images have been
saved.

Task:

Take a single image in JPEG format from your computer.

40 augmented images saved in a folder (e.g., rotated, flipped, zoomed versions of


the original image)

You might also like