The Semi-Automatic Restoration of Regular Paper Fragments

Abstract:

The restoration of regular paper fragments can be solved by edge matching. Edge matching is based primarily on the length and location of the break in writing, however, it is possible that the edges of some regular paper fragments are blank. And so the error of restoration is the inevitable. In this case, it needs manual intervention based on the article content and then eliminate errors. Prior to the establishment of a mathematical model, there is a need for binarization of regular paper fragments with Matlab, then establish vector space model and construct edge contrast matrix, get the order after the Q cluster analysis of fragments. Finally the original picturecan be obtained through splicing and restoring in the order.

• Periodical: Applied Mechanics and Materials (Volume 596)
• Pages: 79-82
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.596.79
• Online since: July 2014
• Authors: Zhong Qiu Ding
• Keywords: Binarization, Edge Contrast Matrix, Manual Intervention, Q Cluster

References

[1] Hun Yu, ZaiKe Hou and XuLi He. Design and implementation of image stitching algorithm based on contour feature[J]. SHIYOU UNIVERSITY(Natural Science Edition), Vol. 2 No. 27 (2003) p.114-p.118 In Chinese.

[2] DongMei Li and YanJie Wang. Image stitching technology based on feature matching[J]. Image processing, Vol. 53 No. 24 (2008) pp.296-298 In Chinese.

[3] QiYuan Jiang, JinXing Xie and Jun Ye. Mathematical Model[M]. High Education Press, 2011 In Chinese.

[4] ZhenYa Zhang, Jin Wang, HongMei Cheng and XvFa Wang. Text space index method based on cosine similarity[J]. Computer Science. No. 09(2005) In Chinese.