Research On High Speed On-line Barcode Quality Examination Technology

Barcode examination is a kind of technology for testing the barcode printing quality,belonging to image processing technology areas. It becomes a hot topic in barcodetechnology that machine vision technology is used for barcode examination andimplemented on an embedded platform. Foreign-related products about barcode examinationhas become better and better. While domestic products, especially the embedded barcodeexamination products, started later so that it is has a gap to compare with foreign products.This thesis mainly studied the barcode quality examination algorithm on DM6437.Around the main line, lucubrated today’s industrial standard of barcode examination,designed a specific program, which is composed of barcode image pre-processing, barcodereading and testing, recognition of the character for human identification, etc. And this thesiscompleted all aspects of the algorithm in the program implemented on DM6437.In the image pre-processing part, according to the actual shooting of the barcode,designed a whole set of algorithms, which is composed of binary, component-labeling, anglemeasurement, interpolation affine. By comparing the pros and cons of these associatedalgorithm, we chose the right solution and achieved an effective extraction of the bar code.By using various optimizing means on DM6437, we optimized the algorithm to improve thespeed of the entire software system. The efficiency of the optimized code is over300timesthan the original code, and the speed of the image processing can be up to7frames persecond.In the part of the barcode examination, according to the examination methods in thenational standards GB/T-18348-2008, eight of the barcode quality-related parameters weremeasured and calculated. The test items were assessed, and the final result were outputted.By searching the array of EAN-13barcode character set, we obtained the decoding value ofthe barcode. The accuracy rate can be up to96%at the first reading.Finally, we used the template matching method to recognize the barcode characters inthe machine recognition. The accuracy rate of character recognition can be up to90.27%.In addition, this thesis also designed the software driver of the button controlling and theresults displaying to achieve the human-machine interface functions.