ABSTRACT
Histogram Equalization (HE) is a popular approach toward improvement of low-contrast images. Many developments have been reported to address several limitations of the conventional HE. However, such algorithms also frequently fail to retain important image characteristics such as brightness, noise characteristics, and textural contents. Optimization of well-formulated objective function may address such problems. In this paper, natural behavior-inspired optimization techniques have been adopted in a hybrid manner, where the search dynamics of Artificial Bee Colony (ABC) is clubbed with the crossover operation of Genetic Algorithm (GA). The hybridization is adopted to avoid the complexity of multi-objective optimization, which is mathematically expensive. The hybrid optimization is employed with the objective function formulated with different image quality metrics conveying different image characteristics. The implementations are made with standard database and found to be potential over conventional techniques in terms of both visual and objective comparisons.
