An artificial bee colony algorithm for image contrast enhancement

Abstract

Image Enhancement is a crucial phase in almost every image processing system. It aims at improving both the visual and the informational quality of distorted images. Histogram Equalization (HE) techniques are the most popular approaches for image enhancement for they succeed in enhancing the image and preserving its main characteristics. However, using exhaustive approaches for histogram equalisation is an algorithmically complex task. These HE techniques also fail in offering good enhancement if not so good parameters are chosen. So, new intelligent approaches, using Artificial Intelligence techniques, have been proposed for image enhancement. In this context, this paper proposes a new Artificial Bee Colony (ABC) algorithm for image contrast enhancement. A grey-level mapping technique and a new image quality measure are used. The algorithm has been tested on some test images, and the comparisons of the obtained results with the genetic algorithm have proven its superiority. Moreover, the proposed algorithm has been extended to colour image enhancement and given very promising results. Further qualitative and statistical comparisons of the proposed ABC to the Cuckoo Search (CS) algorithm are also presented in the paper; not only for the adopted grey-level mapping technique, but also with using another common transformation, generally called the local/global transformation.