基于间距匹配的物流条码定位算法

Logistic barcode location based on spacing matching

针对物流分拣系统中需要对高分辨率图像实时译码的要求,文章提出了基于分块多线扫描和边缘间距匹配的条码定位方法。首先,通过4个方向多根平行的扫描线来采样分块图像;然后统计每条扫描线上相邻边缘的间距,再根据同方向扫描线上边缘间距的匹配程度来定位条码区域;最后通过像素梯度方向的方差来确认条码区域,通过投影和线扫描来修正定位结果。对于分辨率为2 448×2 050的条码图像,当分块大小设为50时,定位耗时26ms,准确率达95.6%。实验结果表明,该算法满足实时物流分拣系统的要求。

For the requirement of real-time barcode location with high resolution images in logistic sorting system,a method based on multiple line scanning and spacing matching in image blocks is proposed.Firstly,image blocks are scanned by parallel lines in four orientations.Then the potential barcode area is located by the matching rate of the spacing between edges in each line.Finally,the barcode area is confirmed by the variance of pixel gradient direction and fixed by projection and line scanning.For barcode images with a resolution of 2 448×2 050,the locating time is 26 ms and the precision rate is 95.6%when the block size is set to 50.The experimental result shows that the algorithm meets the requirements of real-time logistic sorting system.