ABSTRACTS OF ARTICLES OF THE JOURNAL "INFORMATION TECHNOLOGIES".
No. 7. Vol. 24. 2018

DOI: 10.17587/it.24.487-495

A. S. Naumov, Software Engineer, alex.naumov53@mail.ru OOO Electra-N, Velikiy Novgorod, 173003, Russian Federation

Fragmented Image Reconstruction Software Development

Article deals with the technical aspects of the software complex development, which is intended for an automated fragmented image reassembling. This software could be used for a flat objects reconstruction with two-dimensional digital images in forensics, archeology, restoration and archive science. Complex is based on previous author's research and theoretical results. Software architecture features, work logic and program interaction mechanisms of the software are described in the article.
Software complex consist of three separate programs, developed with C# language for the .NET 4.0 platform. Fragment description program performs image preparation and feature description. Match searching program performs matching pairs searching and joint localization. Composition visualizer constructs final reassembled images of the composition.
Software architecture provides separation of the basic logic and algorithms for fragment features description and comparing, so that particular feature set could be changed, expanded or reduced for the current application without basic logic change. Data storage formats for a fragment and match descriptors are based on the XML documents. Intermediate images are stored in the ARGB BMP format files.
Match searching is based on the geometrical and color contours, color palette, texture, other features of the surface and user-defined attributes. Software implements some commonly known feature description methods as well as previously developed by the author, such as contour description method, based on multi-response statistical models, and 3-stage multi-feature match searching algorithm. Examples of ceramic fragments reassembling and some other practical results are presented.
Keywords: software complex, image reconstruction, fragment matching, joint searching, composition, contour, attribute, visualization, XML format

P. 487–495

To the contents