The Logistics Barcode ID Character Recognition Method Based on AKAZE Feature Localization
Jindao Zheng, Kuan Li · 2022 IEEE 10th Joint International Information Technology and Artificial Intelligence Conference (ITAIC) · 2022
With the continuous development of science and technology, the rapid rise of e-commerce, and the continuous influx of the logistics industry into our lives. In the transportation of logistics, the sorting is the mainly work, it needs machines or staff using barcode scanners to identify barcodes to achieve sorting. Meantime, it is necessary to align the barcode of the item manually or mechanically with the barcode scanner to accurately identify it. Undoubtedly, a lot of manpower and material resources are required in the sorting process, so the method of computer vision processing is used to achieve automatic detection and identification of barcode ID will have great significance. The authors propose a logistics barcode ID character recognition method based on local positioning of AKAZE features to realize the detection and identification of logistics barcode ID in a long distance and a large range. First, extracting the AKAZE feature of the barcode in the image, using the matching algorithm to localize the barcode area, and then use the OCR (Optical Character Recognition) method based on deep learning to perform the character recognition of the barcode ID. A large number of experiments have been carried out with the real existing parcel barcodes, in which the accuracy rate reaches 99.6%, the speed reaches 0.35s/frame. Besides, the detection and identification of barcode IDs of different types and different length codes can be realized. The adoption in the future will further promote the feasibility of the logistics and transportation industry.