Identification Of Grain Mildewing With ANN Pattern Recognition Software Based On VB And Matlab

Author

Zhao, De-An ; Pan, Tian-Hong ; Zhu, Jian-Yun ; Wang, Qing

Author_Institution

Jiangsu Univ., Zhenjiang

fYear

2008

fDate

6-8 April 2008

Firstpage

808

Lastpage

811

Abstract

This paper presents a smell analysis system used to identify if grain is mildewed. This system is basically composed of smell sensor array, air circuits system, signal conditioning circuits, data acquisition system and ANN pattern recognition software. VB 6.0 is taken as a platform of development while Matlab 6.5 as a tool of data processing and neural network simulation. For higher recognition accuracy, maximal response point and each point around it among data are taken as eigenvalues and sample eigenvalues are trained with three-layer optimization BP neural network. Lab results show it is effective that software application of neural network pattern recognition based on VB and Matlab in auto analysis system of grain quality smell.

Keywords

agricultural products; backpropagation; data acquisition; eigenvalues and eigenfunctions; electronic noses; neural nets; pattern recognition; signal processing equipment; ANN pattern recognition software; Matlab 6.5; VB 6.0; air circuits system; auto analysis system; data acquisition system; data processing; eigenvalues; grain mildewing; grain quality smell; neural network pattern recognition; neural network simulation; signal conditioning circuits; smell analysis system; smell sensor array; three-layer optimization BP neural network; Artificial neural networks; Circuit simulation; Data acquisition; Data processing; Eigenvalues and eigenfunctions; Neural networks; Pattern recognition; Sensor arrays; Sensor systems; Software quality; ANN; Grain; Matlab; Mixed Programming; Pattern Recognition; VB;

fLanguage

English

Publisher

ieee

Conference_Titel

Networking, Sensing and Control, 2008. ICNSC 2008. IEEE International Conference on

Conference_Location

Sanya

Print_ISBN

978-1-4244-1685-1

Electronic_ISBN

978-1-4244-1686-8

Type

conf

DOI

10.1109/ICNSC.2008.4525326

Filename

4525326