NEW APPROACHES FOR CLASSIFICATION OF HYPERSPECTRAL IMAGES OF CHILI PEPPERS WITH AFLATOXINS

Yıl 2010, Cilt: 12 Sayı: 3, 17 - 33, 01.10.2010

Musa Ataş Yasemin Yardımcı Alptekin Temizel

Öz

Many foods (such as hazelnut, pistachio nut, almond, corn, wheat, dried fig, and chili pepper) are
prone to carcinogenic aflatoxin formation during harvesting, production and storage periods.
Chemical methods are used for detection of aflatoxins give accurate results, but they are slow,
expensive and destructive. In this study, intensity histogram features of hyperspectral images of chili
peppers are extracted under halogen and ultraviolet (UV) illumination source. Salient features are
selected by using connection weights of artificial neural networks and minimum redundancy maximum
relevance techniques. With various topologies of artificial neural networks, effect of data fusion on
classification performance is investigated.

Anahtar Kelimeler

Data fusion, Hyperspectral imaging, Food safety, Machine learning, Aflatoxins

AFLATOKSİNLİ BİBERLERİN HİPERSPEKTRAL GÖRÜNTÜLERİNİN SINIFLANDIRILMASI İÇİN YENİ YAKLAŞIMLAR

Öz

prone to carcinogenic aflatoxin formation during harvesting, production and storage periods. Chemical methods are used for detection of aflatoxins give accurate results, but they are slow, expensive and destructive. In this study, intensity histogram features of hyperspectral images of chili peppers are extracted under halogen and ultraviolet (UV) illumination source. Salient features are selected by using connection weights of artificial neural networks and minimum redundancy maximum relevance techniques. With various topologies of artificial neural networks, effect of data fusion on classification performance is investigated

Anahtar Kelimeler

Veri tümleştirme, Hiperspektral görüntüleme, Gıda güvenliği, Makine öğrenmesi, Aflatoksin

Kaynakça

• Bishop C.M. (1995): “Neural Network for Pattern Recognition”, Oxford University Press Inc., USA.
• Bourlard H., Kamp Y. (1988): “Auto-Association by Multilayer Perceptrons and Singular Value Decomposition”, Bio. Cybernetics, Cilt 59, s. 291-294.
• ElMasrya G., Ning Wangb C. V. (2008): “Detecting Chilling Injury In Red Delicious Apple Using Hyperspectral Imaging And Neural Networks”, Postharvest Biology and Technology, Cilt 52, s. 1-8.
• Hamamoto Y., Uchimura S., Tomita S. (1996): “On The Behavior of Artificial Neural Network Classifiers In High-Dimensional Spaces”, IEEE Trans. on Pattern Anal. and Machine Intel, Cilt 18, No. 5, s. 571-575.
• Hirano S., Okawara N., Narazaki S. (1998): “Near Infra Red Detection of Internally Moldy Nuts”, Bioscience, Biotechnology, and Biochemistry., cilt 62, s. 102-107.
• Jardine D., Peter M. L. (2010): “Black Light Test for Aflatoxin Is QuestionableProcess”, http://www.ksre.ksu.edu/news/sty/2003/blacklight_test082803.htm, 17 Ocak 2010‟da kaydedilmiştir.
• Lipps P.E., Mills D. (2010): “Where to Send Grain Samples for mycotoxin Analysis”, http://www.oardc.ohio-state.edu/ohiofieldcropdisease/wheat/mycotoxin%20text2.htm, 20 Ocak 2010‟da kaydedilmiştir.
• Malek J. E., Alımı A. M., Tourki R. (2000): “Effect of the Feature Vector Size on the Generalization Error: The Case of MLPNN and RBFNN Classifiers”, ICPR.
• Pearson T., Wicklow D., Maghirang E., Xie F., Dowell F. (2001): “Detecting Aflatoxin in Single Corn Kernels by Using Transmittance and Reflectance Spectroscopy”, Transactions of the ASAE, Cilt 44, No. 5, s. 1247-1254.
• Peng H. (2005): “Feature Selection Based on Mutual Information”, IEEE Trans. on Pattern Analysis and Machine Intel, s. 1226-1238.
• Yao H., Hruska Z., Brown R. L, Cleveland T.E. (2006): “Hyperspectral BGYF Imaging of Aflatoxin Contaminated Corn Kernels”, Proc. Of SPIE, Cilt 6381, s. 63810B.
• Zeringue H. J., Shih B. Y. (1998) “Extraction and Separation of the BGYF Material from Aflatoxigenic Aspergillus spp. Infected Cotton Lint by HPLC-UV/FL”, J. Agric. Food Chemistry, Cilt 46, s. 1071-1075.