Cigarette finished product tracing label visual identification method

• Main Authors: CHEN HAITAO, SHI DINGKE, SHU MENG, SUN SHUNKAI, SEO WON-GEUN, XU QI
• Format: Patent
• Language: Chinese; English
• Published: 13.01.2023
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; IMAGE DATA PROCESSING OR GENERATION, IN GENERAL; PHYSICS

The invention discloses a cigarette finished product traceability label visual identification method, and the main design concept of the method is that in a cigarette finished product logistics sorting link, a traceability label initial image on a moving cigarette carton is collected, and the traceability label initial image is preprocessed to obtain an uplink code image of the traceability label; and performing feature extraction on the uplink code image and identifying an uplink code sequence of the traceability label through a CRNN model which is subjected to iterative training based on a traceability label sample set in advance. According to the method, the upward codes of the cigarette cartons in the logistics sorting and moving process are recognized, reliable supplement is provided for the traditional cigarette carton sorting process, the problem that the recognition rate is low due to the fact that traceability labels are not pasted correctly or packaging patterns and other interference are effectivel