节点文献

书籍扫描图像畸变参数自动计算精度的提高

Improving the precision of the auto determined distortion parameters for correction of scanned book image

  • 推荐 CAJ下载
  • PDF下载
  • 不支持迅雷等下载工具,请取消加速工具后下载。

【作者】 申杰张全法庄春意齐永奇

【Author】 SHEN Jie, ZHANG Quan-fa, ZHUANG Chun-yi, QI Yong-qi(School of Physics and Engineering, Zhengzhou University, Zhengzhou 450052, China)

【机构】 郑州大学物理工程学院郑州大学物理工程学院 郑州450052郑州450052

【摘要】 建立了一个数据文件,用于存储一些在不同弯曲状态下书籍扫描图像页边灰度的曲线数据以及对应的多项式系数。利用多项式来描述在某种弯曲状态下像素灰度与页面上的对应点到扫描仪工作平面之间的距离关系。任意弯曲状态下的书籍扫描图像可以在数据文件中搜索到与其最接近的弯曲状态,并利用对应的多项式系数自动求得最佳畸变参数,用于图像的畸变校正实验,获得了较好的校正效果,最大误差由未校正时的63%下降到了4.4%。这使得在无需人工测量或调整畸变参数的条件下可自动校正书籍扫描图像。

【Abstract】 A file is established to save the margin-gray-curves data of scanned book image under different twisted status, and the corresponding coefficients of a multinomial that is supposed to describe the relation between the gray-scale of a pixel and the distance from the pixel’s corresponding point to the scanner’s work plane. For distorted book image under random twisted status, by comparing its margin-gray-curve data with the stored margin-gray-curves data, a group of multinomial coefficients whose corresponding twisted status is proximal to that of the to be corrected image can be selected out to determine automatically a group of distortion parameters that is suitable for distortion correction. By experiment, without manually measure or adjust the distortion parameters, distorted book image can be corrected accurately, maximum error is reduced from 63 percent to 4.4 percent.

  • 【分类号】TP391.41
  • 【被引频次】7
  • 【下载频次】100
节点文献中: 

本文链接的文献网络图示:

本文的引文网络