一种快速响应码图像的分割和校正方法

首先分析快速响应码图像的定位特征模式，提出一种基于一维特征模板匹配及利用图像一维扫描波形快速定位条码区域的方法，并在实现图像方位归一化后，对条码边界进行Hough变换得到边界直线和顶点位置，最后利用控制点变换和双线性插值方法实现条码几何失真校正和大小归一化，得到只包含条码的图像。实验表明这些算法能有效消除图像中无关信息和笔迹污染的干扰，并能校正几何失真。最后得到的条码图像可以进一步由条码符号识别算法和解码算法等进行处理。

Location and Segmentation of Quick Response Code Image

The finder pattern of QR code was analyzed first, and a location method based on one-dimensional pattern match using one-dimensional scanned wave of the bareode was put forward to locate the three finders from background image quickly, and then the rotation angle of the barcode was normalized. As for the distortion of the barcode, Hough transform was used on the edge points, which were scanned from 4 orientations, and then control-point transform and spatial bilinear interpolation arithmetic were conducted on the barcode according to its 4 vertexes consisting a square. The resultant barcode could be processed by decoding algorithm. The algorithms presented in the paper were highlighted by some experimental results, and the results show that the algorithms are able to locate and segment a barcode from complex background, and to correct its distortion.