Particle swarm optimized multi-objective histogram equalization for image enhancement

• We developed multi-objective particle swarm optimization based histogram equalization model for better contrast enhancement and preserving brightness of input images.
• The input image histogram is segmented into two using Otsu's threshold.
• Set of all new optimal weighing constraints are formulated and applied on those histograms before equalization.
• More suitable to be employed in consumer electronics, video frame analysis etc.

Histogram Equalization (HE) is a simple and effective technique for enhancing the contrast of the input image. However, it fails to preserve the brightness while enhancing the contrast due to the abrupt mean shift during the process of equalization. Many HE based methods have been developed to overcome the problem of mean shift. But, they suffered from over-enhancement. In this paper, a multi-objective HE model has been proposed in order to enhance the contrast as well as to preserve the brightness. The central idea of this technique is to first segment the histogram of the input image into two using Otsu's threshold. A set of optimized weighing constraints are formulated and applied on both the sub-images. Then, the sub-images are equalized independently and their union produces the contrast enhanced, brightness preserved output image. Here, Particle Swarm Optimization (PSO) is employed to find the optimal constraints. This technique is proved to have an edge over the other contemporary methods in terms of entropy and contrast improvement index.