Globall Histogram Equalization

This repository contains Python implementation of global histogram equalization for AI6121 - Computer Vision course assignment (NTU MSc in AI programme). I have also written a Medium article explaining the concepts.

Introduction to Histogram Equalization for Digital Image Enhancement

Usage

The histogram equalizer is implemented in histogram equalizer.py. The script to process the test images is CV Assignment1.py. The test images should be placed in jpg/ folder. An output folder jpg/equalized/ must be created, where the equalized images will be written to. To use the equalizer, the following command can be coded:

from histogram equalizer import histogram equalizer

#to use the histogram equalizer
eq_img_array , transform map , cdf , pdf = histogram equalizer(img array)

Input:

img array: 2D NumPy array of values to be equalized, in the range of 0 to 255 (uint8). If the array is multi-channel (eg. RGB image array), each channel will be equalized independently.

Output:

eq img array: NumPy array of the equalized values with the same shape as the input array
transform map : the mapping array to transform the intensity values
cdf : cumulative density function of the input array
pdf : normalized value distribution of the input array

Dependencies:

The scripts depends on the following Python packages:

NumPy
PIL
skimage
matplotlib
pandas

On conda environment, the dependencies can be installed via the provided requirements.txt file with the following command:

pip install −r requirements.txt

Demo

A simple demo is provided in Jupyter notebook, found in the folder /demo_notebook/

hariom1625 / cv_histogram_equalization