A memetic particle gravitation optimization algorithm for solving image segmentation

Significant objects can be obtained from the gray or color images via image segmentation by referring to the clustering problem. Inspired by the search strategy of GSA and the rapid convergence of PSO, this study proposes a memetic particle gravitation optimization (MPGO) algorithm, which is a hybrid approach for performing image segmentation. The proposed MPGO algorithm is verified using six well-known images, including Lena, Baboon, F16, Goldhill, Pepper, and Sailboat. MPGO significantly outperforms k-means, PSO and GSA in measures of the PSNR values.