基于边界像素匹配的碎片拼接问题研究

Research on the Problem of Shredded Document Reconstruction Based on Matching of Pixels for Boundary

  • 摘要: 文中针对碎纸机破碎纸片的双面文字拼接复原问题,以边界像素数据差值最小为优化目标,建立0-1规划模型,并设计了实现拼接复原的算法。与以往人工拼接不同,所设计算法采用分治思想将拼接过程分为横向和纵向,借助Matlab进行机器拼接,并利用边界像素匹配得到机器拼接规则。实验结果表明,该方法可以高效、准确地完成数量庞大的单/双面碎纸片的拼接。

     

    Abstract: Based on minimizing the difference value between boundary pixel data, a 0-1 programming model and corresponding algorithm are established for the doubled-sided shredded document reconstruction problem. The designed algorithm is divided into vertical and horizontal reconstruction. Then utilizing the matching of pixels for boundary rule, we obtained the reconstruction of shredded documents. As it is different from the method of artificial recovery we used before, computer becomes the main labor force with Matlab. It turned out that this method of reconstruction can be accomplished efficiently and precisely.

     

/

返回文章
返回