中北大学学报（自然科学版）2018，Vol.39Issue(2)：177-182,6.DOI:10.3969/j.issn.1673-3193.2018.02.012
基于改进的模糊C均值聚类算法的颗粒种子图像分割方法
Particle Seed Images Segmentation Method Based on the Improved Fuzzy C-means Clustering Algorithm
摘要
Abstract
Adding the penalty term in the obj ective function to represent the neighboring pixel value of traditional FCM algorithm was proposed to solve the noise sensitiveness problems of the crop seed digital image segmentation by applying conventional fuzzy c-means (FCM)clustering algorithm.The spatial in-formation of image was used to improve clustering accuracy.Distance factor of neighboring pixel was used to modify spatial effect.Fuzzy weighting factor was used to improve the obj ective function.The experimental results indicate that the improved FCM algorithm's anti-noise performance is better than traditional FCM algorithm when segmenting crop seeds from background image.The program run time is about half of conventional FCM algorithm.The accuracy rate is increased from 93 % to 99 %,which lays a good foundation for crop seeds counting detection by machine vision system.关键词
颗粒种子图像/图像分割/K-means聚类算法/模糊C均值聚类(FCM)算法Key words
particle seed image/image segmentation/K-means clustering algorithm/FCM algorithm分类
信息技术与安全科学引用本文复制引用
王宇,陈婧,王高..基于改进的模糊C均值聚类算法的颗粒种子图像分割方法[J].中北大学学报（自然科学版）,2018,39(2):177-182,6.基金项目
国家自然科学基金资助项目(61573323) （61573323）
山西省留学人员科研资助项目(2015076) （2015076）
