摘要
水果分级是水果商品化处理的重要一环,为实现苹果自动分级,本文以红富士苹果为测试对象,设计了基于机器视觉的自动分级算法。该算法首先对苹果图像进行灰度化、平滑、增强、分割、边缘提取等预处理;然后综合果径大小、红色着色比及果形圆度三种特征进行自动分级。该方法较单一特征的分级算法更符合实际需求,经实验验证,分级正确率可达90%,具有一定的应用价值。
Fruits grading is an important operational link in commercial treatment of fruit.In order to achieve apple automatic grading,taking the Red Fuji apple as test object,the paper proposed an apple grading algorithm based on machine vision.Firstly,image processing technology was used to denote the original image in order to grayscale,denoise,enhancement,segmentation,contour extraction and other processing.Then,extract the diameter,red coloring rate and roundness value of apple.Finally,the apple can be classified via the Multi-feature.Compared with a single feature classification,the grading algorithm is more effective and suitable.Simulation results show that the accuracy rate of the grading algorithm can up to 90%,with certain application value.
作者
苗艳华
杜青
沈花玉
Miao Yanhua;Du Qing;Shen Huayu(Institute of Information and Automation,Tianjin Tianshi College,Tianjin,301700)
出处
《电子测试》
2019年第1期54-55,58,共3页
Electronic Test
基金
天津天狮学院自然科学研究项目"苹果自动分级算法设计"(K15002)
关键词
机器视觉
苹果
自动分级
machine vision
apple
automatic grading
