This documentation outlines the process for resizing all images in a directory from 1920x1080 resolution to any desired size. 😇

In this example, I’ve used a resolution of 630x340.

The script utilizes Albumentations.ai for both image augmentation and bounding box adjustments.

Below is an explanation of the Python script I’ve created. Feel free to explore the code.

Let's break down the Python script line by line:

1. Installing Necessary Libraries

pip install opencv-python albumentations==1.4.13 tqdm numpy

Make sure to install the albumentations v1.4.13.

2. Importing Necessary Libraries

from tqdm import tqdm

• tqdm is a Python library that adds a progress bar to loops. It provides an easy way to monitor the progress of an iterative process, such as processing a batch of images.

import cv2

• cv2 is the OpenCV library, which is used for image processing tasks like reading, writing, and transforming images.

import albumentations as A

• albumentations is a fast and flexible image augmentation library. Here, it's imported as A for easier reference. It allows for various image transformations, such as resizing, flipping, and more, with specific focus on maintaining annotation integrity for tasks like object detection.

import os

• os is a standard Python library that provides a way to interact with the operating system, such as handling file paths, directories, and environment variables.

3. Defining the read_label_file Function