摘要
由于图像具有模糊性且单阈值分割算法不能满足实际需求,同时考虑单一隶属度函数适应性较差,提出基于标准离差法的模糊散度多阈值图像分割算法。将常用的单阈值隶属度函数推广至多阈值形式,并利用标准离差法计算客观权重,通过线性加权为待分割图像构造新的隶属度函数;推导出多阈值α-型模糊散度作为选取最佳阈值的准则函数;采用粒子群算法优化准则函数以降低多阈值分割算法的运行时间。实验结果表明,该算法可以实现复杂图像多阈值分割,改善分割精度。
Due to the fuzziness of the images and the fact that the single-threshold segmentation algorithm cannot meet the actual demand,taking the poor adaptability of the single membership function into account,we propose a fuzzy divergence multi-threshold image segmentation algorithm based on the standard deviation method.We extended the common single-threshold membership function to the multi-threshold form,and used the standard deviation method to calculate the objective weight.A new membership function was constructed by linear weighting for the image to be segmented.Then,we derived the multi-threshold-type fuzzy divergence as the criterion function for selecting the optimal threshold.Finally,the particle swarm optimization was used to optimize the criterion function to reduce the running time of the multi-threshold segmentation algorithm.The experimental results show that the proposed algorithm can achieve multi-threshold segmentation of complex images,and improve the segmentation accuracy.
作者
杨梦
雷博
史露娜
兰蓉
Yang Meng;Lei Bo;Shi Lu’na;Lan Rong(School of Telecommunications and Information Engineering,Xi’an University of Posts and Telecommunications,Xi’an 710121,Shaanxi,China)
出处
《计算机应用与软件》
北大核心
2020年第5期219-225,共7页
Computer Applications and Software
基金
国家自然科学基金项目(61571361,61671377)
西安邮电大学西邮新星团队项目(xyt2016-01)。
关键词
多阈值图像分割
模糊散度
标准离差法
隶属度函数
粒子群优化算法
Multi-threshold image segmentation
Fuzzy divergence
Standard deviation method
Membership function
Particle swarm optimization algorithm
