A priority based genetic algorithm for limited view tomography

Computed tomography is a noninvasive method to know the internal structure of the objects. It has a wide range of applications, i.e. engineering and medical application. This article presents an adaptive supper parents based GA for limited view tomography. Here, we propose two novel algorithms, namely, single supper based GA and multi supper parents based GA. These algorithms are suitable for engineering as well as the medical application for limited view data or sparse data. This article proposes two novel parental selection strategy. This strategy includes all the advantages of deterministic selection methods and stochastic selection methods. These novel selection methods reduce the loss of diversity using the distribution of the selection opportunity of the entire population members. The proposed algorithm uses adaptive crossover and mutation function that function increase the convergence rate, and the method ensure the consistent performance of the algorithm. Experimental results reveal that the proposed algorithm produces satisfactory results with low computation overhead. The proposed algorithm outperforms with other states of the art algorithm for limited view data or sparse data.