+7 (495) 987 43 74 ext. 3304
Join us -              
Рус   |   Eng

Authors

Dunayev A.

Degree
postgraduate, Department of Information Technologies in Automatized Systems, Belarusian State University of Informatics and Radioelectronics
E-mail
alexander.dunaev.89@gmail.com
Location
Minsk
Articles

Using graph theory in images recognition

The classical problem in computer vision, image processing, and machine vision is that of determining whether or not the image data contains some specific object, feature, or activity. Such problem is also called image recognition problem. An original approach to graph isomorphism problem is defined, which is used in images recognition system. A specificity of the suggested approach is based on the usage of graph structure hashing mechanism. Graph hashing is based on such invariant graph characteristic as the shortest distances between the nodes. An application implementing this approach was developed. Developed application can read an image of symbol into a graph and compare graphs by solving the graphs isomorphism problem. Presented approach was tested on graphs of different types, also on big graphs with more than 100 nodes, and gave a correct answer. A big advantage of this approach is the possibility to store hash-codes in database, since hash-code consists of decimal numbers. It gives an opportunity to create a database of different images, such as faces, symbols, signs. Decimal numbers comparison is very fast, that’s why a quick search can be performed in the database of images. In conclusion, this approach can be successfully used for solving many kinds of image recognition problems.
Read more...