The technique of shredded documentation reconstruction is widely used in different areas such as military affairs and archaeology. This process, however, can hardly be finished merely by human efforts. Therefore, it is important to develop an effective method basing on intelligent algorithms. Designed for effectively solving this problem, this paper attempts to develop an automatic reconstructing method by discomposing the process into three parts, which are tackled with by different algorithms respectively. Firstly, the document pieces, which are in the first column of the original document, are picked out by the blank-area-searching algorithm. Then follows the searching of the adjacent document piece by the help of rightward-eduction algorithm and hence all rows in the original document are reconstructed. Considering the special cases of 'endless loop' as well as 'false searching' at this stage, the corresponding methods of fixed-distance-based clustering analysis and pattern recognition are given. Lastly, by revising the eduction algorithms into upward and downward directions, all rows can be putted back into the original document in the right order. Following these three steps, the shredded original document can be reconstructed automatically.
展开▼
