Image Processing (Color Enhancement)

Abstract

This project aims to enhance the lightness attribute of a color image to provide a better visual performance for people who having color blindness. Color segmentation is first conducted to the input image by k-means clustering. A list of score of the image can be calculated based on the segmentation result and a formula regarding to lightness. The particle swarm optimization algorithm is then used to find out a modification to the original image that will have enhanced lightness values. With the use of above operation, similar colors in RGB space or confusing colors viewed by people who having color blindness can be now distinguishable. Statistic result provided is an addition evidence to determine the performance of the modification other than just visual comparison. The project is implemented in MATLAB and carried out through a user interface created in GUIDE.