Denizanası Arama Optimizasyon Algoritması ile Çok-Odaklı Görüntülerin Birleştirilmesi

Abstract

When obtaining an image of a scene, the lens focuses on objects at a certain distance, and objects at other distances are blurred. This is called the limited depth of field problem. An approach for solving this problem is multi-focus image fusion. A clearer view of the entire scene is obtained by using the multi-focus image fusion method. For this method, at least two images captured at different focuses are combined. Various algorithms have been developed for multi-focus image fusion methods. For multi-focus image fusion, pixel-level block-based methods are commonly used. The block size is a factor that significantly affects the fusion performance. As a result, the block size parameter must be improved. The Jellyfish search optimization algorithm (JSA) is used to propose a block-based multi-focus image fusion approach based on the optimal selection of clearer image blocks from source images. The results of DWTPCA, DCHWT, APCA, PCA, SWTDWT and SWT methods, which are traditional image fusion methods, and ABC (artificial bee colony) and JSA optimization algorithms, which are metaheuristic methods, are compared. In addition, it has been determined that the JSA method has better performance than other traditional methods when compared both visually and quantitatively.