Research on Distillation Process of Chinese Spirits Based on Semantic Segmentation Algorithm

Chinese spirits is a distilled liquor with a long history, and its quality is largely determined by the quality of the distillation process. Therefore, all liquor enterprises regard distillation as the top priority of the whole brewing process. Traditionally, experienced brewers manually judge the quality of the distillation process, which has a huge and unstable workload, and will seriously affect the production efficiency and quality. This paper collects the image datasets of Chinese spirits in different distillation stages, and achieved 86.8% distillation bubble segmentation recognition rate through label datasets and semantic segmentation algorithm, so as to improve the intelligent level of distillation process discrimination, and make a useful exploration to improve the production efficiency and quality of Chinese spirits. This paper fills the gap of domestic related research.