Brightness-Level-Of-an-Image

The following repository calculates the brightness level of an image and outputs a number between 0-10 indicating the brightness of the image.

Algorithm Used:

The raw image is first converted to the LAB color space. In this color space, L represents the lightness value, L = 0 represents the darkest black and L = 100 the brightest white. More about this can be read here.
The algorithm sees certain portions of the image and takes the maximum L value in those portions. It then outputs the brightness as the average of the maximum L values of these different portions. The algorithm doesn't consider the regions with 0 pixel values since a rather bright image with some dark portions can be assigned a pretty low brightness due to those 0 valued pixels.

Code Structure:

• SampleImages folder : It contains the sample images whose brightness is to be computed. If a user needs to compute the brightness level, then he/she can insert/delete the images in this same folder.
• brightness python file : This file contains the function for comuputing the brightness level, it forms the core of this repository.
• main python file : It is the starting point of the repository. Run this file to get the brightness levels.

Installation Steps:

Brightness Level requires Python to run. Install the dependencies and run the main python file.

One can first choose to create a virtual environment. The steps for creating and activating the virtual environment are as follows:

$ conda create --name myenv
$ source activate myenv

Later the libraries can be installed as folows:

$ conda install -c conda-forge opencv
$ pip install matplotlib

The python3.7 interpreter can be used further with any ide/jupyter notebook to run the above code then.

Acknowledgements

This repository inspired me to come up with my own repo!