A Discussion on Automatic Reassembling of Shredded Paper via Similarity Measurement Models
Fanyu Meng, Wen Hui Qian · 2014
Reassembling of shredded paper from a collection of randomly mixed fragments is a common application of computer vision technique.It plays an irreplaceable role in several applied disciplines, such as classical forensics, archaeology, and military intelligence.In this paper we address the reassembly problem using a similarity measurement model applied in data mining.We propose a general process model for automatically analyzing the similarity between the gray value matrices of the scanned images in order to judge if they were adjacent to each other in the original document.Here we take both the Euclid distance and the characteristic values into consideration when it comes to similarity measurement.The implementation results provided demonstrate the validity of the proposed technique.