local-histogram-equalization

CS712 Final Project Evan Hosseini Image processing application for contrast enhancement through local histogram equalization.

Description:

This package contains my final project submission.

Contents:

README.md - this file CS712 Final Project - Histogram Equalization.pptx - Project presentation histogramEqualization.py - Source python file for this project light_bean.jpg - Heavily saturated input image for test dark_bean.jpg - Heavily suppressed input image for test test_pattern.jpg - Test image for the local histogram equalization light_bean_eq.jpg - Result of applying global histogram equalization dark_bean_eq.jpg - Result of applying global histogram equalization test_pattern_local_n_3.jpg - Result of applying local histogram equalization with a neighborhood of size 3 test_pattern_local_n_5.jpg - Result of applying local histogram equalization with a neighborhood of size 5 test_pattern_local_n_25.jpg - Result of applying local histogram equalization with a neighborhood of size 25

Dependencies:

• Python3
  • Numpy
  • Matplotlib
  • Pillow

Usage:

Execute the histogramEqualization.py script w/in a python environment. Specify the input file by giving a -i command switch and optionally specify the name of the output file with a -o switch. To perform local histogram equalization instead of global equalization, pass -n switch specifying the size of the neighborhood to use. This parameter must be an odd number. See below for examples on how the images contained in this package were generated:

Global Equalization

evan@pc ~ $ ./histogramEqualization.py -i light_bean.jpg -o light_bean_eq.jpg

Local Equalization w/ neighborhood of size 3

evan@pc ~ $ ./histogramEqualization.py -i test_pattern.jpg -o test_pattern_local_n_3.jpg -n 3

Notes:

The local histogram equalization may take a while.. Took about 10 minutes for me with the test pattern image running on a quad core i7-3632QM.