基于混沌布谷鸟算法的多阈值灰度图像分割  

Multi-level Gray-scale Image Thresholding Based on Chaotic Cuckoo Search Algorithm

作　　者：孙敏 房明磊 SUN Min;FANG Ming-lei(School of Mathematics and Big Data,Anhui University of Science and Technology,Huainan 232001)

机构地区：[1]安徽理工大学数学与大数据学院,淮南232001

出　　处：《长春理工大学学报（自然科学版）》2020年第1期112-119,共8页Journal of Changchun University of Science and Technology(Natural Science Edition)

基　　金：国家自然科学基金资助项目(11601007)。

摘　　要：阈值法分割图像时,最优阈值选取是否合理对图像分割效果至关重要。标准的布谷鸟算法由于后期存在收敛速度慢,易陷入局部最优等现象,难以准确计算最优分割阈值,因此导致图像分割准确率低。为了提高灰度图像分割的效率和准确率,引入一种基于混沌布谷鸟算法的灰度图像多阈值分割方法。改进的算法利用混沌运动的随机性、遍历性和初值敏感性等优点,对最优鸟窝位置加入由Circle映射产生的混沌扰动策略,有效地防止算法陷入局部最优,加快收敛并提高搜索精度。以最大熵作为目标函数,采用改进的算法对其进行优化,找到分割的最优阈值,实现灰度图像分割。选取2幅经典灰度图像,将所提算法的分割结果与标准的布谷鸟算法、粒子群算法进行对比,以此来说明改进算法的分割质量。实验结果表明,相比于其它两种算法,该改进算法能够快速准确地实现图像分割。The optimal threshold is crucial for image segmentationresults when using the threshold method.It is difficult to find optimal threshold accurately owing to the standard cuckoo search algorithmwhich is easy to trap into the local optima under slow convergence rate and lead toinaccurate segmentationresults.In order to enhance the efficiency and accuracy of gray-scaleimage segmentation,multi-level gray-scale image thresholding based on chaotic cuckoo search algorithmis introduced.Because of the randomness,ergodicity and initial value sensitivity of chaotic motion,the chaotic perturbation strategy generated by Circle mapis applied to the optimal cuckoo’s nest position,which caneffectively prevent the algorithm from falling into local optima andimproveconvergence speed and search accuracy.The chaotic cuckoo search algorithm is used to optimize Kapur’sentropyconsidered as the objective function,then the gray-scale image can be segmented by the obtained optimal thresholds.In thisstudy,two classicalgray-scale images are selected to illustrate the segmentation quality of the proposed algorithm;the experimental results are compared to the standard cuckoo search algorithm,particle swarm optimizationand showthat the improved algorithm couldsegment accuratelyand efficiently.

关 键 词：图像分割 布谷鸟算法 混沌 最大熵 

分 类 号：TP319[自动化与计算机技术—计算机软件与理论]