基于机器视觉的草莓自动分级方法研究  

Research on strawberry automatic classification based on the machine vision

作　　者：杨涛[1] 张云伟[1] 苟爽[1] YANG Tao;ZHANG Yun-wei;GOU Shuang(Faculty of Information Engineering and Automation, Kunming University of Science and Technology, Kunming , Yunnan 650500, Chin)

机构地区：[1]昆明理工大学信息工程与自动化学院,云南昆明650500

出　　处：《食品与机械》2018年第3期146-150,共5页Food and Machinery

基　　金：国家自然科学基金项目(编号:51365019)

摘　　要：针对传统草莓自动分级处理系统利用单一特征衡量草莓品质的不足,提出从成熟度、质量、形状三方面综合评估的方法,以快速有效地完成草莓自动分级处理。该方法先计算分析HSV颜色模型中H参数判断草莓成熟度,利用图像投影面积-质量函数关系对成熟度达标的草莓进行质量测算,再采用K均值聚类法与判别分析相结合对质量达标的草莓进行形状分类,并利用加权法计算草莓质量与形状评级分确定草莓品质等级。试验表明,该方法与人工评级相比准确率达到90%以上。Aiming at the shortcoming of using a single feature to measure the quality of strawberry in traditional strawberry automatic grading system,we proposed a new method to evaluate the quality of strawberry from the aspects of maturity,mass and shape.Firstly,the H parameters in the HSV color model was calculated and analyzed to determine the maturity of strawberries.Secondly,the mass of strawberry was measured by the image projection area-mass function,and then the K-means clustering method and the discriminant analysis were used to obtain the shape classification of strawberry.Finally,the composite score of strawberry mass and shape might be calculated by weighting method to determine strawberry quality grade.Compared to the manual rating,the results showed that the accuracy rate of this new method was more than 90%.

关 键 词：机器视觉 草莓 分级处理 K均值聚类 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]