Cuckoo Search Based Threshold Optimization for Initial Seed Selection in Seeded Region Growing

Image segmentation is the process of grouping pixels into different distinct regions. Seeded Region Growing (SRG) is a method of image segmentation, in which a pixel is used as seed pixel from which the regions start growing. Different seeds provide different results. In this paper we propose a threshold optimization technique using cuckoo search optimization algorithm for initial seed selection. The cuckoo search algorithm is used to find the optimal threshold values which maximize the fitness function.