Grayscale digital halftoning using BPSO-GA combined optimization technique

Grayscale digital halftoning is employed to display multilevel images with bi-level displays e.g. printers, LCD displays, etc. Considering 0 and 1 as black and white, respectively, this can be presented as a binary pattern generation task. This paper presents a combination of binary particle swarm optimization (BPSO) and genetic algorithm (GA) optimization and its application to grayscale digital halftoning. The search dynamics of BPSO have been employed to find best solutions in respect to the individual optimization parameters. The crossover and mutation operations of GA have been performed between the best solutions found by BPSO to obtain the optimized solutions. The optimization has been performed with the test images and the results have been presented pictorially. Objective evaluations have also been performed on the resulted images and compared with the halftones generated using state-of-art techniques.