CSA-FFCM算法在玉米种子芽根长度自动化测定中的应用  

Application of CSA-FFCM Algorithm in the Automatic Determinationof Corn Seed Bud Root Length

作　　者：马启良 胡水星[1] 林冬茂[1] 贾良权 祁亨年 MA Qiliang;HU Shuixing;LIN Dongmao;JIA Liangquan;QI Hengnian(Information Technology Center,Huzhou University,Huzhou 313000,China;School of Information Engineering,Huzhou University,Huzhou 313000,China;Postgraduate School,Huzhou University,Huzhou 313000,China)

机构地区：[1]湖州师范学院信息技术中心,浙江湖州313000 [2]湖州师范学院信息工程学院,浙江湖州313000 [3]湖州师范学院研究生院,浙江湖州313000

出　　处：《湖州师范学院学报》2021年第4期42-49,共8页Journal of Huzhou University

基　　金：湖州师范学院校级科研项目(2020XJKJ15);浙江省重点研发计划项目(2019C02013)。

摘　　要：面对实验室大量的种子实验,研究人员通常采用手工测量法定期对种子的芽根生长长度进行测量,但手工测量方法耗力,且易损伤幼苗,从而造成实验无法进行.为能有效、快捷、准确地掌握种子生长发芽情况,基于CSA-FFCM算法的种子芽根长度分析法,研究玉米种子芽根长度自动测量法.对玉米种子芽根图像进行二值化处理,通过CSA-FFCM算法对灰度图像聚类,得到自适应的分割阈值;在二值图像中识别出每粒玉米种子芽根轮廓区域,并对识别出的玉米种子芽根轮廓进行细化处理,自动统计玉米种子芽根长度.实验结果表明,该方法能够自动搜索分割阈值,且处理速度受图像分辨率影响较小、实时性高;自动测定的玉米芽根长度与PS手工测量长度相比,误差小于5%.此测定方法能为后期的自动化测量提供技术参考.It is high cost and easy to damage seedlings to measure the growth length of seed bud and root by hand.In order to effectively,quickly and accurately grasp the growth of seeds,this paper studies the analysis method of seed bud length based on CSA-FFCM algorithm to realize the automatic measurement of seed bud root length.First,the seed bud root image is binarized,and the grayscale image is clustered by the CSA-FFCM algorithm to obtain an adaptive segmentation threshold;second,each seed bud root contour area is identified in the binary image;Finally,the outline of the identified seed bud root is thinned and automatically count the length of the seed bud and root.Experimental results show that the method is not limited by the resolution of the actual captured image,and Compared with the standard length,the automatically determined bud root length is more than 95%accuracy.This provides technical reference for later automated determination.

关 键 词：玉米种子芽根长度 图像分割 快速模糊C-Means聚类 自适应阈值 自动化测定 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]