摘要
在图像处理、特征提取等技术的基础上,设计了一种水果等级分类系统,提出了一种新型的距离分类器。通过提取出水果的颜色、形状和大小3个最具代表性的外在品质特征,作为距离分类器的输入。仿真结果表明,该方法实时性好,分类准确率高,可满足水果深加工生产的需要。
This paper presents the design of a distance classifier,and builds a fruit level classi-fication system based on image processing and feature extraction.Three representative surface quality conditions for apple obtained from images are used as input to a distance classifier.Simula-tion results show that the system has good real-time performance and high rate of classified accu-racy.It can meet practical needs for deep processing of fruits.
出处
《上海电机学院学报》
2014年第6期343-346,共4页
Journal of Shanghai Dianji University
基金
国家自然科学基金项目资助(61402280)
关键词
距离分类器
图像处理
特征提取
外在品质特征
distance classifier
image processing
feature extraction
surface quality
