Title :
Rapid Shelf-Life Identification Model of Citrus Based on Near Infrared Spectroscopy
Author :
Huijun, Liu ; Xiangfeng, Wu
Author_Institution :
Coll. of Metrol. Technol. & Eng., China Jiliang Univ., Hangzhou
Abstract :
The near-infrared spectroscopy (NIR) was used in modeling of rapid shelf-life identification in citrus in this paper. The spectra of 54 citrus yielded in Huangyan were collected at three different times, and were set of class I, II and III (interval of 10 days), respectively. The principal component analysis was applied in characteristic selection, and ten variables were used as the input of the neural network. The shelf-life identification model of citrus based on principal component analysis and neural network by near-infrared spectroscopy was built. 40 examples were predicated, and the prediction precision was 80%.The result shows that the near-infrared spectroscopy technology can be applied in fast identification of shelf-life of citrus fruits perfectly.
Keywords :
agricultural products; infrared spectroscopy; neural nets; principal component analysis; characteristic selection; citrus; near infrared spectroscopy; neural network; principal component analysis; rapid shelf-life identification model; Algorithm design and analysis; Delta modulation; Food technology; Infrared spectra; Life testing; Neural networks; Predictive models; Principal component analysis; Solids; Spectroscopy;
Conference_Titel :
Knowledge Acquisition and Modeling, 2008. KAM '08. International Symposium on
Conference_Location :
Wuhan
Print_ISBN :
978-0-7695-3488-6
DOI :
10.1109/KAM.2008.50
