Research on Cigarette Carton Rechecking Technology Based on Morphological Processing Algorithms

摘要

This paper puts forward a method for automatic recognition of images on cigarette cartons taking advantage of the uniqueness of the barcode on the carton. It is to solve the mismatch problem between the traditional way of manual rechecking and the high-speed automatic sorting system. Based on the square adaptive morphological structure element algorithm, it deals with monochrome images to reduce the interference caused by blank structures within the barcode to the connected domain, so as to attain multiple candidate connected sub-regions. By screening sub-regions, it positions the barcode area. Then, it locates human-readable characters, and cuts out digital items by projection. At last, barcode digits are recognized, and the recognition is done. Experiment results show that this method is effective in removing the interference caused by characters and images on cigarette cartons as well as by illumination noises.