Image processing for monitoring of the cured tobacco process in a bulk-curing stove

Abstract

A control system based on digital image processing was developed for the tobacco curing process in a bulkcuring barn. A laboratory prototype of the barn was built to test the automation of the process and the results with virginia type tobacco curing are presented. The system records in real-time dry-bulb temperature, air relative humidity, the weight loss of a sample of tobacco leaves and the image of a tobacco leaf. A value for the air humidity ratio of 0.045 kg/kg through the process is obtained as the optimal curing condition. A finite state machine was developed to control the curing condition through the image processing. The HSV color system was used in the image processing; the optimum color for the yellowing stage was obtained with 53.4 for the channel H. In this stage, the time derivative of H close to zero was used as the condition of color stability. The time derivative close to zero for the standard deviation of S was used as uniformity color condition in the color fixing stage and in the leaf drying stage. A good fit between the weight loss and the channel H was obtained; because of this, it was used as the indicator to follow the weight loss in leaves.