Optimal Multilevel Thresholding using Improved Gravitational Search Algorithm for Image Segmentation
Yan Kui Sun, Zhenmin Tang, Jianfeng Lu, Pengzhen Du · 2013
In the view of the characteristics of multilevel threholding selection problems, this paper proposed a new method that Optimal Multilevel Thresholding using Improved Gravitational Search Algorithm for Image Segmentation (MTIGSA). In this paper, two strategies are proposed. One is a new gravity coefficient control strategy in order to avoid redundant computation and improve the convergence speed. Another is a mutation strategy to avoid premature convergence and improve the algorithm precision. Experimental results of both solution quality and computation efficiency illustrate the effectiveness and robustness of MTIGSA.