Bulk Flue-cured Tobacco Barn’s Temperature And Humidity Control System Design Based On Image Processing

Currently, bulk flue-cured barn, instead of traditional tobacco curing barn, has beenlargely used in modern tobacco curing. At the same time, since the automated controlsystem is particularly suited to bulk flue-cured barn, it has also been widely applied tothe environmental control in dry and wet bulb temperature of the bulk flue-cured barn.But it exists problems such as low automation level, low the intensive presence of lowlevels of automation, low level of integration system and poor quality of flue-curedtobacco. It’s related to the development of tobacco industry and even the nationaleconomy to solve the problems above, improve the yield and quality of tobaccoproducts and maximize the liberation of labor.In this paper, the main content is bulk flue-cured barn temperature and humiditycontrol system design based on the image processing of tobacco leaf. In structure, it’sdivided into three parts, which are image feature extraction of flue-cured tobacco, thetemperature and humidity control algorithm of bulk flue-cured barn and the hardwareand software implementation of the temperature and humidity control system.The image feature extraction captures tobacco curing images through imageacquisition system and real-time transfers to the epigenous machine side firstly. Afterthe channel division of the tobacco images in the specified color space and thepreprocessing of grey images median filtering in each channel, it extracts the edges ofthe tobacco leaves, tobacco bars and tobacco vein in the tobacco images by using Sobeland Laplace edge extraction operator. Then, the process is the segmentation of tobaccoleaves and tobacco bars by the application of psychology filter. In the next step, itextracts the color feature of tobacco leaves and tobacco bars in the RGB and HIS colorspace separately and draws the changing curves of their color features of the curingprocess with time and temperature. After this, by analyzing multiple sets of curves, wesearch the corresponding relationship of color characteristics value change and thecuring stage change, find common ground and filter out noise points. At the last, wefuse the multiple color changing curves of tobacco leaves and tobacco bars into anexpert color changing curve as a reference curve of image feature control.The PID control algorithm with fuzzy control algorithm is used as the controlalgorithm for this article. Considering the various factors of the barn during tobaccocuring, we build the physical dynamic mathematical model for the intensive tobacco baking system to describe this complex control object which is the barn, and thus it canprovide a more accurate system model for the fuzzy adaptive PID control algorithm.In this paper, the Siemens S7-200series CPU224XP CN AC/DC/Relayprogrammable logic controller(PLC) serves as the main controller. We select the PT100as the wet and dry bulb temperature sensor and Siemens EM231thermal resistanceexpansion modules as the connection of the temperature sensors to complete thetemperature signal analog-to-digital conversion function. Besides, we achieve the signalinteraction between the control system and epigenous machine by using RS485communication mode. Also, we develop the human-computer interface upon theepigenous machine to convenient the remote monitoring and operating and to improvethe degree of system integration.In this paper, we solve the problems of the low automation level system by usingimage processing algorithm to extract all stages of tobacco curing color characteristicsand to draw the expert color changing curve as a reference curve of image featurecontrol. Secondly, we improve the control precision and quality of tobacco curing byusing the barn-based PID control algorithm combined with fuzzy adaptive algorithm.Thirdly, we improve the degree of system integration by using RS485ascommunication mode and operating the human-computer interface.