Baking Duration Prediction and Management System Based on Image Features

Research on the remaining time of tobacco baking is an important necessity and significance for optimizing tobacco production and processing process, improving tobacco quality and yield, saving resources and costs, and realizing intelligent production. This study devised an innovative Tobacco Leaf Baking Duration Prediction and Management System, capitalizing on the profound capabilities of image processing and machine learning algorithms. The focal point of this system is the precise prediction of the baking duration vital for tobacco leaves, which subsequently affects the color development and aromatic properties of the final producta determinant of quality. Utilizing the ResNet model, renowned for its proficiency in image processing, this study meticulously analyzed the coloration changes during the baking cycle. This rigorous analysis, deployed through the pre-trained model, facilitated the accurate prediction of baking times, tailored to optimize the tobacco leaves' quality. To further enhance the practical application of the system, this study incorporated a dataset embellished with varied tobacco leaf images collected at strategic intervals within the baking timeline. This dataset was pivotal in training and validating the model, ensuring its adaptability and accuracy in real-world scenarios. To complement the front-end strengths of the system, this study integrated a user-friendly web-based graphical interface.