Cigarette Review Method Based on Recognition of One-Dimensional Bar Code

The traditional artificial cigarette review methods are difficult to match the high-speed automatic cigarette sorting system. In this paper, a new cigarette recognition method using the uniqueness of cigarette bar codes was proposed. Firstly, a binary image filtering algorithm based on square adaptive structure element of mathematical morphology (SASEMM) was used to get several potential sub-regions, which could minimize bar-space´s interference in marking connected areas. Then the article located bar code area based on sub-region screening method. Finally after locating and segmenting readable characters based on projection method, it recognize bar codes digit and complicate cigarette recognition. The results show that the methods in this paper can get rid of the interference from uneven illumination, cigarette surface´s characters and patterns.