关键词: 图像分割, 阈值分割, K均值

Abstract:

The cuckoo search algorithm (CS) is a bionic algorithm,but its application to image segmentation suffers from such drawbacks as large amount of calculation,appearing local minimum and slow convergence.In order to solve these problems,we propose a multi-threshold image segmentation algorithm based on modified cuckoo search algorithm.This algorithm employs the Otsu method as the fitness function,combines the cuckoo search algorithm with the K-means algorithm for better diversity of population,and adaptively determines the number and range of the thresholds to find the optimal thresholds of the image to be segmented.Experimental results show that the MCS algorithm outperforms the K-means and the cuckoo search (CS) in terms of segmentation thresholds and segmentation effect.

Key words: image segmentation;threshold segmentation;K means