P.O. Arkhipov - Ph. D. (Eng.), Senior Research Scientist, Orel Branch of Institute of Informatics Problems of RAS. E-mail: arpaul@mail.ru

In this article one of the tasks of processing and analysis of color images is discussed - the problem of automatic detection of a plurality of features of the color images. Under the features of images are usually understood well distinguishable points on the images. We propose a technology that allows us to divide the process in video analysis on two sets of activities: analysis of immutable characteristics - work with local characteristics and analysis of the characteristics associated with instability of the color (changing light of the same items during the day). The purpose of this work is the creation of information technology of diagnosing abnormal changes in the structure of surfaces in the color image extracted from an arbitrary video streams to detect the changed local features and color differences. To achieve this it is necessary to solve the following problems: split the video stream into many key-frames of a particular format; diagnosis of abnormal changes in the structure of surfaces key-frames. The implementation of the described technology will allow tracking of abnormal changes in the structure of the images in the video surveillance system, promptly sending a notice of the penetration of the alarming device and the operator of the surveillance system.

 

1. Konushin A.C. Vvedenie v kompjuternoe zrenie [EHlektronnyjj resurs] // Lektorium. 20.02.12.  URL: https://www.lektorium.tv/lecture/13541-id=13541.
2. Arkhipov P.O., Sorokin A.I. Informacionnaja tekhnologija formirovanija kharakteristicheskikh priznakov predvaritelnojj identifikacii cvetnykh izobrazhenijj // Informacionnye sistemy i tekhnologii. Orel: Gosuniversitet UNPK. 2012. № 6/74. S. 36−44.
3. Duda R., Hart P. Pattern Classification and Scene Analysis. John Wiley and Sons. 1973. P. 271−272.
4. Arkhipov P.O. Algoritmicheskaja model formirovanija metadannykh cvetnykh izobrazhenijj [EHlektronnyjj resurs] // ITNOP: materialy VIMezhdunar. nauchno-tekhnich. konf. 22−23.05.2014. Orel: Gosuniversitet UNPK. 2014. URL: http://irsit.ru/files/article/392.pdf. 4 s.
5. Rosten E., Drummond T. Fusing Points and Lines for High Performance Tracking. 2005.
6. Rosten E., Drummond T. Machine learning for high-speed corner detection // European Conference on Computer Vision 1. P. 430−443. Doi: 10.1007/11744023_34. 2006.
7. Lowe D.G. Object recognition from local scale-invariant features // Proc. of the International Conference on Computer Vision 2. P. 1150−1157. Doi:10.1109/ICCV.1999.790410.
8. Lowe D.G. Distinctive Image Features from Scale-Invariant Keypoints // International Journal of Computer Vision. 2004. V. 60. № 2. P. 91−110.
9. Arkhipov O.P., Manjakov JU.A., Sirotinin D.O.Metod generacii virtualnojj setki opornykh tochek na cvetnykh izobrazhenijakh [EHlektronnyjj resurs] // ITNOP: materialy VMezhdunar. nauchno-tekhnich. konf. 17−18.05.2012. Orel: Gosuniversitet UNPK. 2012. URL: http://irsit.ru/article/149.pdf. 6 s.
10. Arkhipov O.P., Manjakov JU.A., Sirotinin D.O. Informacionnaja model metoda pofragmentnogo analiza i predstavlenija izmenenijj naturnogo obekta // Informacionnye sistemy i tekhnologii. Orel: Gosuniversitet UNPK. 2012. № 1(69). S. 17−25.