A novel intelligent control system for flue-curing barns based on real-time image features

• Novel intelligent control to set dry and wet bulb temperatures for curing barns.
• Image features of tobacco leaves extracted in real time from vision system.
• ANN model to identify the setpoint values of the dry and wet bulb temperatures.
• Simulations and experimental results show effectiveness of control system.

Most intensive tobacco curing systems are manually operated requiring the curers to frequently observe the status of tobacco leaves and in order to achieve the desired temperature and relative humidity, curers adjust the setpoint values of dry and wet bulb temperatures and the time to change to the next setpoints. Control is therefore subjective and it is difficult to maintain consistent high quality curing. A novel intelligent control system based on the real-time image processing of the tobacco leaves images to monitor the status of the tobacco leaves was developed. A neural network based approach was designed to identify the setpoints for the dry and wet bulb temperatures, and the time to change to the next setpoints. Inputs were 12 extracted image features obtained from an image processing algorithm and the measured dry and wet bulb temperatures in the barn. Without any manual intervention by curers, the developed intelligent control system achieved real-time monitoring and management of the curing process. The effectiveness of the developed intelligent control system was demonstrated by simulation and experiment.