基于BP神经网络二维码区域提取

BP neural network based on the two-dimensional code region extraction

在复杂背景下的二维码区域定位一直是QR Code二维条码解码过程中的难题之一。二维码区域扫描定位是通过二维码的图形特点来实现,其存在扫描定位效率较低的缺点。为此提出在扫描定位之前通过图像处理结合BP神经网络实现QR Code二维码条码区域提取方法。火车票通过图像预处理得到可能是二维码的区域块,提取经图像处理后的二维码区域块图像特征并结合BP神经网络过滤出正确的二维码区域。此方法实现了寻找一幅图像中二维码区域的图像,结合二维码图形扫描定位方法,提高了二维码扫描定位的效率,得到了较好的效果。

A two-dimensional code location in complex background is always one of the difficult problems of QR Code two- dimensional barcode decoding process. The two-dimensional code region scanning positioning is realized by the graphical features of two-dimensional codes. Its disadvantage is low efficiency of scanning positioning. This paper combines BP neural network to realize QR Code two-dimensional bar code region extraction method by processing the image in the scanning position before. Train tickets are probably block two-dimensional code by image preproeessing, feature is extracted by regional image two-dimensional code image processing and BP neural network filtering out the two-dimensional code area in the right. This method realizes the image for a two- dimensional code image regions, combining with the two-dimensional code image scanning positioning method, improves the efficiency of the two-dimensional code scanning and positioning, and good effect is obtained.