Application of artificial neural network and adaptive neuro-fuzzy inference systems in determining the moisture content in green tea sheets based on colored parameters

Shahabi Ghoyonloe, Mohammad; Raﬁee, Shahin; Mohtasebi, Seyed Saeid; Hosseinpour, Soleiman

doi:10.22059/ijbse.2014.50120

Application of artificial neural network and adaptive neuro-fuzzy inference systems in determining the moisture content in green tea sheets based on colored parameters

Document Type : Research Paper

Authors

Mohammad Shahabi Ghoyonloe ¹

Shahin Raﬁee ²

Seyed Saeid Mohtasebi ²

Soleiman Hosseinpour ³

¹ MS.c. Student

² Professor, University College of Agriculture and Natural Resources, University of Tehran

³ former Graduated Student, University College of Agriculture and Natural Resources, University of Tehran

10.22059/ijbse.2014.50120

Abstract

Using image processing and artificial intelligence systems in agriculture and food industry is increasing daily. The purpose of this research is to study the feasibility of using image processing technique in predicting process of moisture content changes on green tea sheets during the drying using predictive artificial intelligence systems such as: artificial neural networks and adaptive neuro-fuzzy inference system. The drying experiments were conducted at five temperatures of 50, 60, 70, 80 and 90 °C and three air flow rates of 0.5, 1 and 1.5 m/s using thin layer method. The results gained out of extracting colorful images took from upper view of samples were applied as input data of artificial intelligence systems for determining moisture content. Finally, the best results predicted by the artificial neural network with two hidden layers contained (12 neurons in the first layer and 15 neurons on the second layer) with correlation coefficient of 0.948 and root mean square error of 0.092, respectively.

Keywords

image processing

Artificial Intelligence

Drying

green tea