Barcode recognition device and method based on artificial intelligence semantic segmentation

Abstract

The invention relates to a barcode recognition device and method based on artificial intelligence semantic segmentation. Coarse positioning is performed on a bar code image in a complex environment through a convolutional neural network semantic segmentation algorithm to obtain bar code roughing information, then fine positioning is performed on the roughing information to obtain a decoded image and corresponding decoding additional information, and finally decoding operation is performed on the decoded image according to the decoding additional information to obtain bar code electronic information. According to the invention, bar codes under complex conditions can be identified, and the method has the characteristics of high efficiency, high speed and high robustness.

Description

technical field [0001] The invention relates to the field of barcode recognition, in particular to a barcode recognition device and method based on artificial intelligence semantic segmentation. Background technique [0002] The task of barcode recognition is mainly composed of three parts: detection, location and decoding. The role of the detection task is to identify objects with barcode characteristics within the entire image range; the role of the positioning task is to give the specific coordinates of the barcode on the basis of detection (x1, y1, x2, y2 define a rectangular frame) ; The role of the decoding task is to parse out the electronic data information of the located barcode. [0003] The front-end physical information collection of barcode recognition has two methods based on laser and image sensor. Laser-based barcode recognition is only effective for a single barcode. When multiple barcodes appear, it is difficult for this technology to realize simultaneous...

AI Technical Summary

Problems solved by technology

Therefore, it does not work well in complex environments, complex backgrounds, etc., especially at long distances, with high noise, etc.

In particular, when there are multiple codes in the same picture (a picture contains 2 or more barcodes), due to the traversal search characteristics of traditional technology in detection and positioning, the increase in the number of barcodes will further lead to the difficulty of detection and positioning. big, slow

[0006] In short, the existing barcode recognition methods perform poorly in terms of barcode recognition speed and accuracy in situations such as multiple barcodes, complex backgrounds, and complex environments.

Images