Prediction of output moisture content of dill from hot-air conveyor belt dryer using machine vision

Predicting the output moisture content of product from the conveyor belt hot air dryer for controlling the drying process is one of important parameters. Therefore, in this research, a conveyor belt dryer with a hot air flow equipped with a machine vision system was developed. Dryer also consists of air temperature and conveyor belt speed controlling section, lighting and imaging system. The control sections for air temperature and conveyor belt speed include SSR relays and a programmed algorithm in MATLAB software environment. The machine vision section comprises three cameras placed at the beginning, middle, and end of the conveyor belt. In this study, experiments were conducted at two temperature levels of 50 and 60 °C and three levels of conveyor belt speed for each treatment. Then, using the developed image processing algorithm in MATLAB, the changes in shrinkage were extracted and analyzed. Finally, the out moisture content of the product from dryer was modeled using the ANN. The results of this study indicated that the out moisture content and shrinkage of the dried product are dependent on temperature of dryer and speed of the conveyor belt. Specifically, with an increase in temperature and a decrease in conveyor belt speed, the degree of shrinkage increases Finally, results revealed that the ANN with 4-12-1 structure had best prediction performance with 1.06e-6, 1.24e-6, 9.46e-7 of RMSE and 0.9999, 0.9999, 0.9999 of R, respectively for training, validation and testing data.