基于模糊聚类的蓝莓成熟分级无损检测应用研究  

Research on Non-destructive Testing of Blueberry Ripening Grading Based on Fuzzy Clustering

作　　者：王龙 刘绪川 史兴强 梅英杰 WANG Long;LIU Xuchuan;SHI Xingqiang;MEI Yingjie(China Key System&Integrated Circuit Co.,Ltd.,Wuxi 214072;School of Instrument Science and Engineering,Harbin Institute of Technology,Harbin 150001)

机构地区：[1]中科芯集成电路有限公司,无锡214072 [2]哈尔滨工业大学仪器科学与工程学院,哈尔滨150001

出　　处：《计算机与数字工程》2024年第2期548-553,共6页Computer & Digital Engineering

基　　金：江苏省科技成果转化高性能实时微处理器系列产品研发及产业化项目(编号:BA2019012)资助。

摘　　要：由于现有的水果成熟分级技术通用性不够,不能对蓝莓成熟进行有效分级。因此提出基于模糊聚类的蓝莓成熟分级方法。通过图像处理技术分别对蓝莓的面积、周长、圆形度和着色分布进行计算,最后确定蓝莓成熟隶属函数和分级标准。试验表明,基于模糊聚类的蓝莓成熟分级方法能够有效地、快速地对蓝莓进行分级,其准确率为90%。Due to the insufficient versatility of existing fruit ripening and grading techniques,it is not possible to effectively grade ripe blueberries.Therefore,a blueberry ripe grading method based on fuzzy clustering is proposed.The area,perimeter,roundness and coloring distribution of blueberries are calculated by image processing technology,and finally the blueberry ripe membership function and grading standard are determined.Experiments show that the blueberry ripe grading method based on fuzzy clustering can effectively and quickly grade blueberries,with an accuracy rate of 90%.

关 键 词：成熟分级 模糊聚类 圆形度 着色分布 隶属函数 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]