In order to promote the long-distance online trade, splitting goods near their origin, and pricing in accordance with the quality of potted flowers, we developed a quality image classification system for pot poinsettia. The image processing tools, the object recognition tool – YOLO V5, and other tools are used to establish characteristics identification parameters and class intervals of the quantity, flower roundness, cyathium compactness, flower stem symmetry and plant height and width of pot poinsettia. Based on the classification standards, the overall class and important information are created and calculated to provide buyers with references.