www.old.international-agrophysics.org / zeszyty


International Agrophysics
wydawca:Instytut Agrofizyki
im. B. Dobrzańskiego
PAN
w Lublinie
ISSN: 0236-8722


vol. 24, nr. 4 (2010)

poprzedni artykuł wróć do listy artykułów następny artykuł
Feasibility of near infrared spectroscopy for analysis of date fruits
(pobierz wersję PDF)
Mireei S.A.1, Mohtasebi S.S.1, Massudi R.2, Rafiee S. 1, and Arabanian A.S.2
1 Department of Agricultural Machinery Engineering, Faculty of Agricultural Engineering and Technology, University of Tehran, P.O. Box 4111, Karaj 31587-77871, Iran
2 Laser and Plasma Research Institute, Shahid Beheshti University, G.C., Evin, Tehran 1983963113, Iran

vol. 24 (2010), nr. 4, pp. 351-356
streszczenie This paper deals with a research to study a near infrared spectroscopy as a nondestructive method for discrimination of date fruits according to four main ripening stages. For this purpose, a near infrared spectra of dates were acquired in the range of 900-1700 nm. Principal component analysis was performed to reduce the dimensionality of the spectral data and then the first five principal components were used as the inputs to a back propagation artificial neural network (BP-ANN) to classify the dates based on the ripening stages. The principal component analysis – artificial neural network (PCA-ANN) model provided satisfactory discrimination results with an accuracy of 95.5% for test sample set. Moreover, partial least square method with different data preprocessing methods was applied to predict the moisture content and soluble solids content of date fruits. The best predictive models showed the coefficient of determination values of 0.98 and 0. 0.96 with residual predictive deviation values of 6.22 and 5.03 for the moisture and the soluble solids content, respectively. A near infrared spectroscopy appeared to be a good method for both classification of Shahani date fruits according to ripening stages and also determination of their maturity indices.
słowa kluczowe NIR spectroscopy, date fruit, artificial neural network, discrimination, moisture content