自适应步长下多阈值彩色图像的全局分割方法  

Global Segmentation Method for Multi-threshold Color Images under Adaptive Step Size

作　　者：鲁秋菊[1] 拓守恒[1] LU Qiuju;TUO Shouheng(School of Mathematics and Computer Science,Shaanxi University ofTechnology,Hanzhong 723000,Shaanxi Province,China)

机构地区：[1]陕西理工大学数学与计算机科学学院,陕西汉中723000

出　　处：《吉林大学学报（理学版）》2019年第1期82-88,共7页Journal of Jilin University:Science Edition

基　　金：国家自然科学基金(批准号:61301237);陕西省教育厅专项科研基金(批准号:15JK1134;16JK1157);陕西理工大学科研项目(批准号:SLGKY16-14)

摘　　要：针对传统多阈值彩色图像分割方法将步长设为小于距离参数的定值,有时会因步长过大而越过最优结果的问题,提出一种自适应步长下多阈值彩色图像全局分割方法.首先,对彩色图像进行预处理,在不降低彩色图像质量的前提下缩减颜色总数,以提高分割效率;然后,根据多阈值彩色图像全阈值分割目标函数,将混沌优化理论与粒子群优化算法相结合,通过混沌粒子群优化算法对多阈值彩色图像全局分割目标函数进行求解;最后,结合自适应步长下多阈值彩色图像全局分割方法,得到最优彩色图像阈值分割结果.实验结果表明,该方法的分割效果、精度、稳定性和收敛性均较好.Aiming at the problem that the step size of traditional multi-threshold color image segmentation method was less than the fixed value of distance parameter,sometimes the step size was too large to exceed the optimal result,we proposed global segmentation method for multi-threshold color images under adaptive step size.Firstly,the color image was preprocessed to solve the problem of reducing the total number of colors without reducing the quality of the color image in order to improve the segmentation efficiency.Secondly,according to the objective function of multi-threshold color image segmentation,the chaotic optimization theory was combined with particle swarm optimization algorithm,the global segmentation objective function of multi-threshold color image was solved by chaotic particle swarm optimization algorithm.Finally,combined with the global segmentation method for multi-threshold color image under adaptive step size,the optimal color image threshold segmentation results were obtained.The experimental results show that the segmentation effect,accuracy,stability and convergence of this method are good.

关 键 词：自适应步长 多阈值 彩色图像 全局 分割 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]