基于凸包算法的二维条码定位

Research on two-dimensional bar code positioning approach based on convex hull algorithm

在二维条码识别过程中,只有精确定位条码区域在原始图像中的位置,才能进一步做后续的矫正处理.文中以二维条码Data Matrix为例将计算几何中的凸包概念应用于条码定位,从像素点序列中筛选出凸包顶点以减少待处理像素点的数量.同时利用Data Matrix的定位图形特征设定筛选条件,最终得到用于定位条码区域的凸包顶点.凸包算法克服了传统直线检测算法因全局检测而需要大量存储空间的缺点.实验结果表明此算法能对扭曲或倾斜的二维条码区域进行有效的定位,并减少精确定位消耗的处理时间.

During the process of recognizing two-dimensional bar code, it is important for the next correcting process that the area of bar code is positioned accurately in the original image. The concept of convex hull in computational geometry is used in locating bar code through in case of Data Matrix in the paper. The convex hull points are filtered from the list of pixels in order to reduce the number of the pixels processed. At the same time, the selection conditions are set by the locating image features in Data Matrix. Finally the convex hull points used in locating bar code are gotten. This algorithm overcomes the weakness in traditional linear detection algorithms, in which the large storage is needed in order to support global detection. The result shows that the algorithm can locate two-dimensional bar code effectively even under the distorted or tilted situation. It also reduces the processing time in accurate position.