Title :
The Research of Feature Extraction Methods in the Tomatoes Detection
Author :
Wang Xiao-liang ; Guan Xiao-Hui ; Wang Jian-yuan ; Yang Yong-qing ; Liu Hai-yan
Author_Institution :
Office of Acad. Affairs, Northeast Dianli Univ., Changchun, China
Abstract :
Feature selection can reduce the feature space dimension and improve recognition. In the discriminatory fresh degree of tomatoes by electronic nose, it used three kinds of feature extraction methods: sheath coefficient characteristics, similitude entropy characteristics and the energy characteristics, and the three methods were compared respectively. The results showed that similitude entropy characteristics feature extraction methods has its advantages in electronic nose detection.
Keywords :
agricultural products; electronic noses; entropy; feature extraction; electronic nose; electronic nose detection; entropy characteristics; feature extraction methods; feature selection; tomatoes detection; Actuators; Chemicals; Electronic noses; Entropy; Feature extraction; Gas detectors; Vectors; electronic nose; feature extraction; sensor array;
Conference_Titel :
Instrumentation, Measurement, Computer, Communication and Control (IMCCC), 2012 Second International Conference on
Conference_Location :
Harbin
Print_ISBN :
978-1-4673-5034-1
DOI :
10.1109/IMCCC.2012.170
