摘要
人工分级在双孢蘑菇工厂化生产中具有标准不统一、生产效率低和劳动量大等问题,提出新鲜双孢蘑菇采收和自动化分级方法。采用机械采收器采收双孢蘑菇,采集双孢蘑菇图像。在图像预处理阶段中,利用最大阈值分割法和全局阈值分割法对双孢蘑菇图像进行第一次分水岭处理,去除双孢蘑菇图像中存在的阴影,通过闭运算和Canny算子对双孢蘑菇图像进行第二次分水岭处理,去除图像中的非边缘部分,消除图像中存在的阴影和非边缘部分对分级造成的干扰。提取双孢蘑菇的大小特征和形状特征,将提取到的特征输入支持向量机分类函数中,实现新鲜双孢蘑菇的自动化分级。结果表明:所提方法的特征提取精准度高、分级效率高。
There are some problems in the industrial production of Agaricus bisporus,such as inconsistent standards,low production efficiency and large labor quantity.A.bisporus images were collected by mechanical harvester.In the image preprocessing stage,using threshold segmentation method and the largest global threshold segmentation method to first divide the A.bisporus image,get rid of A.bisporus images that exist in the shadows,by closing operation and canny operator of A.bisporus second watershed image,and get rid of the edges of the image,eliminate shadows that exist in the image and the edge part cause interference to the sizing.The size and shape features of A.bisporus were extracted,and the extracted features were input into the classification function of support vector machine to realize automatic classification of fresh A.bisporus.The experimental results show that the proposed method has high feature extraction accuracy and high classification efficiency.
作者
刘韦
LIU Wei(University of Jinan Quancheng College,Penglai 265600,China)
出处
《中国食用菌》
北大核心
2019年第10期56-59,63,共5页
Edible Fungi of China
基金
山东省教育科学“十三五”规划专项课题(BYZN201902).
关键词
双孢蘑菇
图像预处理
自动化分级
特征提取
Agaricus bisporus
image preprocessing
automated grading
feature extraction
