Application of Feature Extraction through Convolution Neural Networks and SVM Classifier for Robust Grading of Apples

作     者：Yuan CAI Clarence W.DE SILVA Bing LI Liqun WANG Ziwen WANG 

作者机构：Harbin Institute of Technology at ShenzhenShenzhen 518000 The University of British ColumbiaVancouver BCCanada V6T 1Z4 

出 版 物：《Instrumentation》 (仪器仪表学报（英文版）)

年 卷 期：2019年第6卷第4期

页      面：59-71页

学科分类：0832[工学-食品科学与工程（可授工学、农学学位）] 12[管理学] 1201[管理学-管理科学与工程(可授管理学、工学学位)] 08[工学] 081104[工学-模式识别与智能系统] 0835[工学-软件工程] 0811[工学-控制科学与工程] 083203[工学-农产品加工及贮藏工程] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

主　　题：Apple Grading k-nearest Neighbour Method Convolutional Neural Network Support Vector Machine Machine Learning 

摘      要：This paper proposes a novel grading method of apples,in an automated grading device that uses convolutional neural networks to extract the size,color,texture,and roundness of an *** developed machine learning method uses the ability of learning representative features by means of a convolutional neural network(CNN),to determine suitable features of apples for the grading *** information is fed into a one-to-one classifier that uses a support vector machine(SVM),instead of the softmax output layer of the *** this manner,Yantai apples with similar shapes and low discrimination are graded using four different *** fusion model using both CNN and SVM classifiers is much more accurate than the simple k-nearest neighbor(KNN),SVM,and CNN model when used separately for grading,and the learning ability and the generalization ability of the model is correspondingly increased by the combined *** tests are carried out using the automated grading device that is developed in the present *** is verified that the actual effect of apple grading using the combined CNN-SVM model is fast and accurate,which greatly reduces the manpower and labor costs of manual grading,and has important commercial prospects.